Delay nix-gc until /home is mounted

Prevents starting the garbage collector before the remote FS are
mounted, in particular /home. Otherwise, all the gcroots which have
symlinks in /home will be considered stale and they will be removed.

See: #79
Reviewed-by: Aleix Roca Nonell <aleix.rocanonell@bsc.es>

doc/install.md

@@ -150,3 +150,27 @@ And update grub.
 ```
 # nix build .#nixosConfigurations.xeon02.config.system.build.kexecTree -v
 ```
+
+## Chain NixOS in same disk with other systems
+
+To install NixOS on a partition along another system which controls the GRUB,
+first disable the grub device, so the GRUB is not installed in the disk by
+NixOS (only the /boot files will be generated):
+
+```
+boot.loader.grub.device = "nodev";
+```
+
+Then add the following entry to the old GRUB configuration:
+
+```
+menuentry 'NixOS' {
+        insmod chain
+        search --no-floppy --label nixos --set root
+        configfile /boot/grub/grub.cfg
+}
+```
+
+The partition with NixOS must have the label "nixos" for it to be found. New
+system configuration entries will be stored in the GRUB configuration managed
+by NixOS, so there is no need to change the old GRUB settings.

flake.lock

@@ -6,14 +6,15 @@
         "home-manager": "home-manager",
         "nixpkgs": [
           "nixpkgs"
-        ]
+        ],
+        "systems": "systems"
       },
       "locked": {
-        "lastModified": 1690228878,
-        "narHash": "sha256-9Xe7JV0krp4RJC9W9W9WutZVlw6BlHTFMiUP/k48LQY=",
+        "lastModified": 1720546205,
+        "narHash": "sha256-boCXsjYVxDviyzoEyAk624600f3ZBo/DKtUdvMTpbGY=",
         "owner": "ryantm",
         "repo": "agenix",
-        "rev": "d8c973fd228949736dedf61b7f8cc1ece3236792",
+        "rev": "de96bd907d5fbc3b14fc33ad37d1b9a3cb15edc6",
         "type": "github"
       },
       "original": {
@@ -23,18 +24,23 @@
       }
     },
     "bscpkgs": {
+      "inputs": {
+        "nixpkgs": [
+          "nixpkgs"
+        ]
+      },
       "locked": {
-        "lastModified": 1690560045,
-        "narHash": "sha256-39ZP+FIzlWoN3c43hReBYpStg4RLYw/z7TdxCQmOvTM=",
+        "lastModified": 1713974364,
+        "narHash": "sha256-ilZTVWSaNP1ibhQIIRXE+q9Lj2XOH+F9W3Co4QyY1eU=",
         "ref": "refs/heads/master",
-        "rev": "b4a20d7c3af854b39682484adfd1c7979319f439",
-        "revCount": 841,
+        "rev": "de89197a4a7b162db7df9d41c9d07759d87c5709",
+        "revCount": 937,
         "type": "git",
-        "url": "https://pm.bsc.es/gitlab/rarias/bscpkgs.git"
+        "url": "https://git.sr.ht/~rodarima/bscpkgs"
       },
       "original": {
         "type": "git",
-        "url": "https://pm.bsc.es/gitlab/rarias/bscpkgs.git"
+        "url": "https://git.sr.ht/~rodarima/bscpkgs"
       }
     },
     "darwin": {
@@ -45,11 +51,11 @@
         ]
       },
       "locked": {
-        "lastModified": 1673295039,
-        "narHash": "sha256-AsdYgE8/GPwcelGgrntlijMg4t3hLFJFCRF3tL5WVjA=",
+        "lastModified": 1700795494,
+        "narHash": "sha256-gzGLZSiOhf155FW7262kdHo2YDeugp3VuIFb4/GGng0=",
         "owner": "lnl7",
         "repo": "nix-darwin",
-        "rev": "87b9d090ad39b25b2400029c64825fc2a8868943",
+        "rev": "4b9b83d5a92e8c1fbfd8eb27eda375908c11ec4d",
         "type": "github"
       },
       "original": {
@@ -67,11 +73,11 @@
         ]
       },
       "locked": {
-        "lastModified": 1682203081,
-        "narHash": "sha256-kRL4ejWDhi0zph/FpebFYhzqlOBrk0Pl3dzGEKSAlEw=",
+        "lastModified": 1703113217,
+        "narHash": "sha256-7ulcXOk63TIT2lVDSExj7XzFx09LpdSAPtvgtM7yQPE=",
         "owner": "nix-community",
         "repo": "home-manager",
-        "rev": "32d3e39c491e2f91152c84f8ad8b003420eab0a1",
+        "rev": "3bfaacf46133c037bb356193bd2f1765d9dc82c1",
         "type": "github"
       },
       "original": {
@@ -82,11 +88,11 @@
     },
     "nixpkgs": {
       "locked": {
-        "lastModified": 1692447944,
-        "narHash": "sha256-fkJGNjEmTPvqBs215EQU4r9ivecV5Qge5cF/QDLVn3U=",
+        "lastModified": 1720957393,
+        "narHash": "sha256-oedh2RwpjEa+TNxhg5Je9Ch6d3W1NKi7DbRO1ziHemA=",
         "owner": "NixOS",
         "repo": "nixpkgs",
-        "rev": "d680ded26da5cf104dd2735a51e88d2d8f487b4d",
+        "rev": "693bc46d169f5af9c992095736e82c3488bf7dbb",
         "type": "github"
       },
       "original": {
@@ -102,6 +108,21 @@
         "bscpkgs": "bscpkgs",
         "nixpkgs": "nixpkgs"
       }
+    },
+    "systems": {
+      "locked": {
+        "lastModified": 1681028828,
+        "narHash": "sha256-Vy1rq5AaRuLzOxct8nz4T6wlgyUR7zLU309k9mBC768=",
+        "owner": "nix-systems",
+        "repo": "default",
+        "rev": "da67096a3b9bf56a91d16901293e51ba5b49a27e",
+        "type": "github"
+      },
+      "original": {
+        "owner": "nix-systems",
+        "repo": "default",
+        "type": "github"
+      }
     }
   },
   "root": "root",

flake.nix

@@ -3,7 +3,8 @@
     nixpkgs.url = "github:NixOS/nixpkgs/nixos-unstable";
     agenix.url = "github:ryantm/agenix";
     agenix.inputs.nixpkgs.follows = "nixpkgs";
-    bscpkgs.url = "git+https://pm.bsc.es/gitlab/rarias/bscpkgs.git";
+    bscpkgs.url = "git+https://git.sr.ht/~rodarima/bscpkgs";
+    bscpkgs.inputs.nixpkgs.follows = "nixpkgs";
   };
 
   outputs = { self, nixpkgs, agenix, bscpkgs, ... }:
@@ -16,15 +17,19 @@ let
 in
   {
     nixosConfigurations = {
-      hut   = mkConf "hut";
-      owl1  = mkConf "owl1";
-      owl2  = mkConf "owl2";
-      eudy  = mkConf "eudy";
-      koro  = mkConf "koro";
-      bay   = mkConf "bay";
-      lake2 = mkConf "lake2";
+      hut     = mkConf "hut";
+      owl1    = mkConf "owl1";
+      owl2    = mkConf "owl2";
+      eudy    = mkConf "eudy";
+      koro    = mkConf "koro";
+      bay     = mkConf "bay";
+      lake2   = mkConf "lake2";
+      raccoon = mkConf "raccoon";
     };
 
-    packages.x86_64-linux.hut = self.nixosConfigurations.hut.pkgs;
+    packages.x86_64-linux = self.nixosConfigurations.hut.pkgs // {
+      bscpkgs = bscpkgs.packages.x86_64-linux;
+      nixpkgs = nixpkgs.legacyPackages.x86_64-linux;
+    };
   };
 }

keys.nix

@@ -0,0 +1,29 @@
+# As agenix needs to parse the secrets from a standalone .nix file, we describe
+# here all the public keys
+rec {
+  hosts = {
+    hut   = "ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAICO7jIp6JRnRWTMDsTB/aiaICJCl4x8qmKMPSs4lCqP1 hut";
+    owl1  = "ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIMqMEXO0ApVsBA6yjmb0xP2kWyoPDIWxBB0Q3+QbHVhv owl1";
+    owl2  = "ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIHurEYpQzNHqWYF6B9Pd7W8UPgF3BxEg0BvSbsA7BAdK owl2";
+    eudy  = "ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIL+WYPRRvZupqLAG0USKmd/juEPmisyyJaP8hAgYwXsG eudy";
+    koro  = "ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIImiTFDbxyUYPumvm8C4mEnHfuvtBY1H8undtd6oDd67 koro";
+    bay   = "ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAICvGBzpRQKuQYHdlUQeAk6jmdbkrhmdLwTBqf3el7IgU bay";
+    lake2 = "ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAINo66//S1yatpQHE/BuYD/Gfq64TY7ZN5XOGXmNchiO0 lake2";
+  };
+
+  hostGroup = with hosts; rec {
+    compute    = [ owl1 owl2 ];
+    playground = [ eudy koro ];
+    storage    = [ bay lake2 ];
+    monitor    = [ hut ];
+
+    system     = storage ++ monitor;
+    safe       = system ++ compute;
+    all        = safe ++ playground;
+  };
+
+  admins = {
+    rarias = "ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIE1oZTPtlEXdGt0Ak+upeCIiBdaDQtcmuWoTUCVuSVIR rarias@hut";
+    root   = "ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIII/1TNArcwA6D47mgW4TArwlxQRpwmIGiZDysah40Gb root@hut";
+  };
+}

m/bay/configuration.nix

@@ -2,21 +2,21 @@
 
 {
   imports = [
-    ../common/main.nix
-    ../common/monitoring.nix
+    ../common/xeon.nix
+    ../module/monitoring.nix
   ];
 
   # Select the this using the ID to avoid mismatches
   boot.loader.grub.device = "/dev/disk/by-id/wwn-0x55cd2e414d53562d";
 
+  boot.kernel.sysctl = {
+    "kernel.yama.ptrace_scope" = lib.mkForce "1";
+  };
+
   environment.systemPackages = with pkgs; [
     ceph
   ];
 
-  services.slurm = {
-    client.enable = lib.mkForce false;
-  };
-
   networking = {
     hostName = "bay";
     interfaces.eno1.ipv4.addresses = [ {
@@ -27,16 +27,25 @@
       address = "10.0.42.40";
       prefixLength = 24;
     } ];
+    firewall = {
+      extraCommands = ''
+        # Accept all incoming TCP traffic from lake2
+        iptables -A nixos-fw -p tcp -s lake2 -j nixos-fw-accept
+        # Accept monitoring requests from hut
+        iptables -A nixos-fw -p tcp -s hut -m multiport --dport 9283,9002 -j nixos-fw-accept
+        # Accept all Ceph traffic from the local network
+        iptables -A nixos-fw -p tcp -s 10.0.40.0/24 -m multiport --dport 3300,6789,6800:7568 -j nixos-fw-accept
+      '';
+    };
   };
 
   services.ceph = {
     enable = true;
     global = {
       fsid = "9c8d06e0-485f-4aaf-b16b-06d6daf1232b";
-      monHost = "10.0.42.40";
-      monInitialMembers = "10.0.42.40";
-      publicNetwork = "10.0.42.40/24";
-      clusterNetwork = "10.0.42.40/24";
+      monHost = "10.0.40.40";
+      monInitialMembers = "bay";
+      clusterNetwork = "10.0.40.40/24"; # Use Ethernet only
     };
     extraConfig = {
       # Only log to stderr so it appears in the journal
@@ -50,7 +59,7 @@
       enable = true;
       daemons = [ "mds0" "mds1" ];
       extraConfig = {
-        "host" = "10.0.42.40";
+        "host" = "bay";
       };
     };
     mgr = {

m/common/base.nix

@@ -0,0 +1,20 @@
+{
+  # All machines should include this profile.
+  # Includes the basic configuration for an Intel server.
+  imports = [
+    ./base/agenix.nix
+    ./base/august-shutdown.nix
+    ./base/boot.nix
+    ./base/env.nix
+    ./base/fs.nix
+    ./base/hw.nix
+    ./base/net.nix
+    ./base/nix.nix
+    ./base/ntp.nix
+    ./base/rev.nix
+    ./base/ssh.nix
+    ./base/users.nix
+    ./base/watchdog.nix
+    ./base/zsh.nix
+  ];
+}

m/common/base/agenix.nix

@@ -0,0 +1,9 @@
+{ agenix, ... }:
+
+{
+  imports = [ agenix.nixosModules.default ];
+
+  environment.systemPackages = [
+    agenix.packages.x86_64-linux.default
+  ];
+}

m/common/base/august-shutdown.nix

@@ -0,0 +1,14 @@
+{
+  # Shutdown all machines on August 2nd at 11:00 AM, so we can protect the
+  # hardware from spurious electrical peaks on the yearly electrical cut for
+  # manteinance that starts on August 4th.
+  systemd.timers.august-shutdown = {
+    description = "Shutdown on August 2nd for maintenance";
+    wantedBy = [ "timers.target" ];
+    timerConfig = {
+      OnCalendar = "*-08-02 11:00:00";
+      RandomizedDelaySec = "10min";
+      Unit = "systemd-poweroff.service";
+    };
+  };
+}

m/common/base/boot.nix

@@ -2,7 +2,7 @@
 
 {
   # Use the GRUB 2 boot loader.
-  boot.loader.grub.enable = lib.mkForce true;
+  boot.loader.grub.enable = true;
 
   # Enable GRUB2 serial console
   boot.loader.grub.extraConfig = ''
@@ -11,14 +11,12 @@
     terminal_output --append serial
   '';
 
-  # Enable serial console
-  boot.kernelParams = [
-    "console=tty1"
-    "console=ttyS0,115200"
-  ];
-
   boot.kernel.sysctl = {
     "kernel.perf_event_paranoid" = lib.mkDefault "-1";
+
+    # Allow ptracing (i.e. attach with GDB) any process of the same user, see:
+    # https://www.kernel.org/doc/Documentation/security/Yama.txt
+    "kernel.yama.ptrace_scope" = "0";
   };
 
   boot.kernelPackages = pkgs.linuxPackages_latest;

m/common/base/env.nix

@@ -0,0 +1,35 @@
+{ pkgs, config, ... }:
+
+{
+  environment.systemPackages = with pkgs; [
+    vim wget git htop tmux pciutils tcpdump ripgrep nix-index nixos-option
+    nix-diff ipmitool freeipmi ethtool lm_sensors ix cmake gnumake file tree
+    ncdu config.boot.kernelPackages.perf ldns
+    # From bsckgs overlay
+    osumb
+  ];
+
+  programs.direnv.enable = true;
+
+  # Increase limits
+  security.pam.loginLimits = [
+    {
+      domain = "*";
+      type = "-";
+      item = "memlock";
+      value = "1048576"; # 1 GiB of mem locked
+    }
+  ];
+
+  environment.variables = {
+    EDITOR = "vim";
+    VISUAL = "vim";
+  };
+
+  programs.bash.promptInit = ''
+    PS1="\h\\$ "
+  '';
+
+  time.timeZone = "Europe/Madrid";
+  i18n.defaultLocale = "en_DK.UTF-8";
+}

m/common/base/fs.nix

@@ -6,20 +6,19 @@
       fsType = "ext4";
     };
 
+  # Trim unused blocks weekly
+  services.fstrim.enable = true;
+
   swapDevices =
     [ { device = "/dev/disk/by-label/swap"; }
     ];
 
-  # Mount the home via NFS
-  fileSystems."/home" = {
-    device = "10.0.40.30:/home";
-    fsType = "nfs";
-    options = [ "nfsvers=3" "rsize=1024" "wsize=1024" "cto" "nofail" ];
-  };
-
   # Tracing
   fileSystems."/sys/kernel/tracing" = {
     device = "none";
     fsType = "tracefs";
   };
+
+  # Mount a tmpfs into /tmp
+  boot.tmp.useTmpfs = true;
 }

m/common/base/net.nix

@@ -0,0 +1,19 @@
+{ pkgs, ... }:
+
+{
+  networking = {
+    enableIPv6 = false;
+    useDHCP = false;
+
+    firewall = {
+      enable = true;
+      allowedTCPPorts = [ 22 ];
+    };
+
+    hosts = {
+      "84.88.53.236" = [ "ssfhead.bsc.es" "ssfhead" ];
+      "84.88.51.152" = [ "raccoon" ];
+      "84.88.51.142" = [ "raccoon-ipmi" ];
+    };
+  };
+}

m/common/base/nix.nix

@@ -0,0 +1,57 @@
+{ pkgs, nixpkgs, bscpkgs, theFlake,  ... }:
+
+{
+  nixpkgs.overlays = [
+    bscpkgs.bscOverlay
+    (import ../../../pkgs/overlay.nix)
+  ];
+
+  nix = {
+    nixPath = [
+      "nixpkgs=${nixpkgs}"
+      "jungle=${theFlake.outPath}"
+    ];
+
+    registry = {
+      nixpkgs.flake = nixpkgs;
+      jungle.flake = theFlake;
+    };
+
+    settings = {
+      experimental-features = [ "nix-command" "flakes" ];
+      sandbox = "relaxed";
+      trusted-users = [ "@wheel" ];
+      flake-registry = pkgs.writeText "global-registry.json"
+        ''{"flakes":[],"version":2}'';
+    };
+
+    gc = {
+      automatic = true;
+      dates = "weekly";
+      options = "--delete-older-than 30d";
+    };
+  };
+
+  # The nix-gc.service can begin its execution *before* /home is mounted,
+  # causing it to remove all gcroots considering them as stale, as it cannot
+  # access the symlink. To prevent this problem, we force the service to wait
+  # until /home is mounted as well as other remote FS like /ceph.
+  systemd.services.nix-gc = {
+    # Start remote-fs.target if not already being started and fail if it fails
+    # to start. It will also be stopped if the remote-fs.target fails after
+    # starting successfully.
+    bindsTo = [ "remote-fs.target" ];
+    # Wait until remote-fs.target fully starts before starting this one.
+    after = [ "remote-fs.target"];
+    # Ensure we can access a remote path inside /home
+    unitConfig.ConditionPathExists = "/home/Computational";
+  };
+
+  # This value determines the NixOS release from which the default
+  # settings for stateful data, like file locations and database versions
+  # on your system were taken. It‘s perfectly fine and recommended to leave
+  # this value at the release version of the first install of this system.
+  # Before changing this value read the documentation for this option
+  # (e.g. man configuration.nix or on https://nixos.org/nixos/options.html).
+  system.stateVersion = "22.11"; # Did you read the comment?
+}

m/common/base/rev.nix

@@ -0,0 +1,21 @@
+{ theFlake, ... }:
+
+let
+  # Prevent building a configuration without revision
+  rev = if theFlake ? rev then theFlake.rev
+    else throw ("Refusing to build from a dirty Git tree!");
+in {
+  # Save the commit of the config in /etc/configrev
+  environment.etc.configrev.text = rev + "\n";
+
+  # Keep a log with the config over time
+  system.activationScripts.configRevLog.text = ''
+    BOOTED=$(cat /run/booted-system/etc/configrev 2>/dev/null || echo unknown)
+    CURRENT=$(cat /run/current-system/etc/configrev 2>/dev/null || echo unknown)
+    NEXT=${rev}
+    DATENOW=$(date --iso-8601=seconds)
+    echo "$DATENOW booted=$BOOTED current=$CURRENT next=$NEXT" >> /var/configrev.log
+  '';
+
+  system.configurationRevision = rev;
+}

m/common/base/ssh.nix

@@ -0,0 +1,22 @@
+{ lib, ... }:
+
+let
+  keys = import ../../../keys.nix;
+  hostsKeys = lib.mapAttrs (name: value: { publicKey = value; }) keys.hosts;
+in
+{
+  # Enable the OpenSSH daemon.
+  services.openssh.enable = true;
+
+  # Connect to intranet git hosts via proxy
+  programs.ssh.extraConfig = ''
+    Host bscpm02.bsc.es bscpm03.bsc.es gitlab-internal.bsc.es alya.gitlab.bsc.es
+      User git
+      ProxyCommand nc -X connect -x hut:23080 %h %p
+  '';
+
+  programs.ssh.knownHosts = hostsKeys // {
+    "gitlab-internal.bsc.es".publicKey = "ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIF9arsAOSRB06hdy71oTvJHG2Mg8zfebADxpvc37lZo3";
+    "bscpm03.bsc.es".publicKey = "ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIM2NuSUPsEhqz1j5b4Gqd+MWFnRqyqY57+xMvBUqHYUS";
+  };
+}

m/common/base/users.nix

@@ -1,6 +1,10 @@
-{ ... }:
+{ pkgs, ... }:
 
 {
+  imports = [
+    ../../module/jungle-users.nix
+  ];
+
   users = {
     mutableUsers = false;
     users = {
@@ -26,6 +30,7 @@
           "ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAINa0tvnNgwkc5xOwd6xTtaIdFi5jv0j2FrE7jl5MTLoE ram@mio"
           "ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIGYcXIxe0poOEGLpk8NjiRozls7fMRX0N3j3Ar94U+Gl rarias@hal"
         ];
+        shell = pkgs.zsh;
       };
 
       arocanon = {
@@ -41,18 +46,73 @@
           "ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIGdphWxLAEekicZ/WBrvP7phMyxKSSuLAZBovNX+hZXQ aleix@kerneland"
         ];
       };
+    };
 
+    jungleUsers = {
       rpenacob = {
         uid = 2761;
         isNormalUser = true;
         home = "/home/Computational/rpenacob";
         description = "Raúl Peñacoba";
         group = "Computational";
+        hosts = [ "owl1" "owl2" "hut" ];
         hashedPassword = "$6$TZm3bDIFyPrMhj1E$uEDXoYYd1z2Wd5mMPfh3DZAjP7ztVjJ4ezIcn82C0ImqafPA.AnTmcVftHEzLB3tbe2O4SxDyPSDEQgJ4GOtj/";
         openssh.authorizedKeys.keys = [
           "ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIFYfXg37mauGeurqsLpedgA2XQ9d4Nm0ZGo/hI1f7wwH rpenacob@bsc"
         ];
       };
+
+      anavarro = {
+        uid = 1037;
+        isNormalUser = true;
+        home = "/home/Computational/anavarro";
+        description = "Antoni Navarro";
+        group = "Computational";
+        hosts = [ "hut" "raccoon" ];
+        hashedPassword = "$6$QdNDsuLehoZTYZlb$CDhCouYDPrhoiB7/seu7RF.Gqg4zMQz0n5sA4U1KDgHaZOxy2as9pbIGeF8tOHJKRoZajk5GiaZv0rZMn7Oq31";
+        openssh.authorizedKeys.keys = [
+          "ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAILWjRSlKgzBPZQhIeEtk6Lvws2XNcYwHcwPv4osSgst5 anavarro@ssfhead"
+        ];
+      };
+
+      abonerib = {
+        uid = 4541;
+        isNormalUser = true;
+        home = "/home/Computational/abonerib";
+        description = "Aleix Boné";
+        group = "Computational";
+        hosts = [ "owl1" "owl2" "hut" "raccoon" ];
+        hashedPassword = "$6$V1EQWJr474whv7XJ$OfJ0wueM2l.dgiJiiah0Tip9ITcJ7S7qDvtSycsiQ43QBFyP4lU0e0HaXWps85nqB4TypttYR4hNLoz3bz662/";
+        openssh.authorizedKeys.keys = [
+          "ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIIIFiqXqt88VuUfyANkZyLJNiuroIITaGlOOTMhVDKjf abonerib@bsc"
+        ];
+      };
+
+      vlopez = {
+        uid = 4334;
+        isNormalUser = true;
+        home = "/home/Computational/vlopez";
+        description = "Victor López";
+        group = "Computational";
+        hosts = [ "koro" ];
+        hashedPassword = "$6$0ZBkgIYE/renVqtt$1uWlJsb0FEezRVNoETTzZMx4X2SvWiOsKvi0ppWCRqI66S6TqMBXBdP4fcQyvRRBt0e4Z7opZIvvITBsEtO0f0";
+        openssh.authorizedKeys.keys = [
+          "ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIGMwlUZRf9jfG666Qa5Sb+KtEhXqkiMlBV2su3x/dXHq victor@arch"
+        ];
+      };
+
+      dbautist = {
+        uid = 5649;
+        isNormalUser = true;
+        home = "/home/Computational/dbautist";
+        description = "Dylan Bautista Cases";
+        group = "Computational";
+        hosts = [ "hut" ];
+        hashedPassword = "$6$a2lpzMRVkG9nSgIm$12G6.ka0sFX1YimqJkBAjbvhRKZ.Hl090B27pdbnQOW0wzyxVWySWhyDDCILjQELky.HKYl9gqOeVXW49nW7q/";
+        openssh.authorizedKeys.keys = [
+          "ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIAb+EQBoS98zrCwnGKkHKwMLdYABMTqv7q9E0+T0QmkS dbautist@bsc-848818791"
+        ];
+      };
     };
 
     groups = {

m/common/base/zsh.nix

@@ -0,0 +1,91 @@
+{ pkgs, ... }:
+
+{
+  environment.systemPackages = with pkgs; [
+    zsh-completions
+    nix-zsh-completions
+  ];
+
+  programs.zsh = {
+    enable = true;
+    histSize = 1000000;
+
+    shellInit = ''
+      # Disable new user prompt
+      if [ ! -e ~/.zshrc ]; then
+        touch ~/.zshrc
+      fi
+    '';
+
+    promptInit = ''
+      # Note that to manually override this in ~/.zshrc you should run `prompt off`
+      # before setting your PS1 and etc. Otherwise this will likely to interact with
+      # your ~/.zshrc configuration in unexpected ways as the default prompt sets
+      # a lot of different prompt variables.
+      autoload -U promptinit && promptinit && prompt default && setopt prompt_sp
+    '';
+
+    # Taken from Ulli Kehrle config:
+    # https://git.hrnz.li/Ulli/nixos/src/commit/2e203b8d8d671f4e3ced0f1744a51d5c6ee19846/profiles/shell.nix#L199-L205
+    interactiveShellInit = ''
+      source "${pkgs.zsh-history-substring-search}/share/zsh-history-substring-search/zsh-history-substring-search.zsh"
+
+      # Save history immediately, but only load it when the shell starts
+      setopt inc_append_history
+
+      # dircolors doesn't support alacritty:
+      # https://lists.gnu.org/archive/html/bug-coreutils/2019-05/msg00029.html
+      export LS_COLORS='rs=0:di=01;34:ln=01;36:mh=00:pi=40;33:so=01;35:do=01;35:bd=40;33;01:cd=40;33;01:or=40;31;01:mi=00:su=37;41:sg=30;43:ca=00:tw=30;42:ow=34;42:st=37;44:ex=01;32:*.tar=01;31:*.tgz=01;31:*.arc=01;31:*.arj=01;31:*.taz=01;31:*.lha=01;31:*.lz4=01;31:*.lzh=01;31:*.lzma=01;31:*.tlz=01;31:*.txz=01;31:*.tzo=01;31:*.t7z=01;31:*.zip=01;31:*.z=01;31:*.dz=01;31:*.gz=01;31:*.lrz=01;31:*.lz=01;31:*.lzo=01;31:*.xz=01;31:*.zst=01;31:*.tzst=01;31:*.bz2=01;31:*.bz=01;31:*.tbz=01;31:*.tbz2=01;31:*.tz=01;31:*.deb=01;31:*.rpm=01;31:*.jar=01;31:*.war=01;31:*.ear=01;31:*.sar=01;31:*.rar=01;31:*.alz=01;31:*.ace=01;31:*.zoo=01;31:*.cpio=01;31:*.7z=01;31:*.rz=01;31:*.cab=01;31:*.wim=01;31:*.swm=01;31:*.dwm=01;31:*.esd=01;31:*.avif=01;35:*.jpg=01;35:*.jpeg=01;35:*.mjpg=01;35:*.mjpeg=01;35:*.gif=01;35:*.bmp=01;35:*.pbm=01;35:*.pgm=01;35:*.ppm=01;35:*.tga=01;35:*.xbm=01;35:*.xpm=01;35:*.tif=01;35:*.tiff=01;35:*.png=01;35:*.svg=01;35:*.svgz=01;35:*.mng=01;35:*.pcx=01;35:*.mov=01;35:*.mpg=01;35:*.mpeg=01;35:*.m2v=01;35:*.mkv=01;35:*.webm=01;35:*.webp=01;35:*.ogm=01;35:*.mp4=01;35:*.m4v=01;35:*.mp4v=01;35:*.vob=01;35:*.qt=01;35:*.nuv=01;35:*.wmv=01;35:*.asf=01;35:*.rm=01;35:*.rmvb=01;35:*.flc=01;35:*.avi=01;35:*.fli=01;35:*.flv=01;35:*.gl=01;35:*.dl=01;35:*.xcf=01;35:*.xwd=01;35:*.yuv=01;35:*.cgm=01;35:*.emf=01;35:*.ogv=01;35:*.ogx=01;35:*.aac=00;36:*.au=00;36:*.flac=00;36:*.m4a=00;36:*.mid=00;36:*.midi=00;36:*.mka=00;36:*.mp3=00;36:*.mpc=00;36:*.ogg=00;36:*.ra=00;36:*.wav=00;36:*.oga=00;36:*.opus=00;36:*.spx=00;36:*.xspf=00;36:*~=00;90:*#=00;90:*.bak=00;90:*.old=00;90:*.orig=00;90:*.part=00;90:*.rej=00;90:*.swp=00;90:*.tmp=00;90:*.dpkg-dist=00;90:*.dpkg-old=00;90:*.ucf-dist=00;90:*.ucf-new=00;90:*.ucf-old=00;90:*.rpmnew=00;90:*.rpmorig=00;90:*.rpmsave=00;90:';
+
+      # From Arch Linux and GRML
+      bindkey "^R" history-incremental-pattern-search-backward
+      bindkey "^S" history-incremental-pattern-search-forward
+
+      # Auto rehash for new binaries
+      zstyle ':completion:*' rehash true
+      # show a nice menu with the matches
+      zstyle ':completion:*' menu yes select
+
+      bindkey '^[OA' history-substring-search-up   # Up
+      bindkey '^[[A' history-substring-search-up   # Up
+
+      bindkey '^[OB' history-substring-search-down # Down
+      bindkey '^[[B' history-substring-search-down # Down
+
+      bindkey '\e[1~' beginning-of-line            # Home
+      bindkey '\e[7~' beginning-of-line            # Home
+      bindkey '\e[H'  beginning-of-line            # Home
+      bindkey '\eOH'  beginning-of-line            # Home
+
+      bindkey '\e[4~' end-of-line                  # End
+      bindkey '\e[8~' end-of-line                  # End
+      bindkey '\e[F ' end-of-line                  # End
+      bindkey '\eOF'  end-of-line                  # End
+
+      bindkey '^?'    backward-delete-char         # Backspace
+      bindkey '\e[3~' delete-char                  # Del
+      # bindkey '\e[3;5~' delete-char                # sometimes Del, sometimes C-Del
+      bindkey '\e[2~' overwrite-mode               # Ins
+
+      bindkey '^H'      backward-kill-word         # C-Backspace
+
+      bindkey '5~'      kill-word                  # C-Del
+      bindkey '^[[3;5~' kill-word                  # C-Del
+      bindkey '^[[3^'   kill-word                  # C-Del
+
+      bindkey "^[[1;5H" backward-kill-line         # C-Home
+      bindkey "^[[7^"   backward-kill-line         # C-Home
+
+      bindkey "^[[1;5F" kill-line                  # C-End
+      bindkey "^[[8^"   kill-line                  # C-End
+
+      bindkey '^[[1;5C' forward-word               # C-Right
+      bindkey '^[0c'    forward-word               # C-Right
+      bindkey '^[[5C'   forward-word               # C-Right
+
+      bindkey '^[[1;5D' backward-word              # C-Left
+      bindkey '^[0d'    backward-word              # C-Left
+      bindkey '^[[5D'   backward-word              # C-Left
+    '';
+  };
+}

m/common/main.nix

@@ -1,94 +0,0 @@
-{ config, pkgs, nixpkgs, bscpkgs, agenix, theFlake, ... }:
-
-{
-  imports = [
-    ./boot.nix
-    ./fs.nix
-    ./hw.nix
-    ./net.nix
-    ./ntp.nix
-    ./slurm.nix
-    ./ssh.nix
-    ./users.nix
-    ./watchdog.nix
-  ];
-
-  nixpkgs.overlays = [
-    bscpkgs.bscOverlay
-    (import ../../pkgs/overlay.nix)
-  ];
-
-  nix.nixPath = [
-    "nixpkgs=${nixpkgs}"
-    "bscpkgs=${bscpkgs}"
-    "jungle=${theFlake.outPath}"
-  ];
-
-  nix.registry.nixpkgs.flake = nixpkgs;
-  nix.registry.bscpkgs.flake = bscpkgs;
-  nix.registry.jungle.flake = theFlake;
-
-  system.configurationRevision =
-    if theFlake ? rev
-    then theFlake.rev
-    else throw ("Refusing to build from a dirty Git tree!");
-
-  environment.systemPackages = with pkgs; [
-    vim wget git htop tmux pciutils tcpdump ripgrep nix-index nixos-option
-    nix-diff ipmitool freeipmi ethtool lm_sensors ix cmake gnumake file tree
-    ncdu config.boot.kernelPackages.perf ldns
-    # From bsckgs overlay
-    bsc.osumb
-  ];
-
-  systemd.services."serial-getty@ttyS0" = {
-    enable = true;
-    wantedBy = [ "getty.target" ];
-    serviceConfig.Restart = "always";
-  };
-
-  # Increase limits
-  security.pam.loginLimits = [
-    {
-      domain = "*";
-      type = "-";
-      item = "memlock";
-      value = "1048576"; # 1 GiB of mem locked
-    }
-  ];
-
-  time.timeZone = "Europe/Madrid";
-  i18n.defaultLocale = "en_DK.UTF-8";
-
-  environment.variables = {
-    EDITOR = "vim";
-    VISUAL = "vim";
-  };
-
-  nix.settings.experimental-features = [ "nix-command" "flakes" ];
-  nix.settings.sandbox = "relaxed";
-  nix.settings.trusted-users = [ "@wheel" ];
-  nix.gc.automatic = true;
-  nix.gc.dates = "weekly";
-  nix.gc.options = "--delete-older-than 30d";
-
-  programs.zsh.enable = true;
-  programs.zsh.histSize = 100000;
-
-  programs.bash.promptInit = ''
-    PS1="\h\\$ "
-  '';
-
-  # Copy the NixOS configuration file and link it from the resulting system
-  # (/run/current-system/configuration.nix). This is useful in case you
-  # accidentally delete configuration.nix.
-  #system.copySystemConfiguration = true;
-
-  # This value determines the NixOS release from which the default
-  # settings for stateful data, like file locations and database versions
-  # on your system were taken. It‘s perfectly fine and recommended to leave
-  # this value at the release version of the first install of this system.
-  # Before changing this value read the documentation for this option
-  # (e.g. man configuration.nix or on https://nixos.org/nixos/options.html).
-  system.stateVersion = "22.11"; # Did you read the comment?
-}

m/common/slurm.nix

@@ -1,42 +0,0 @@
-{ lib, ... }:
-
-{
-  systemd.services.slurmd.serviceConfig = {
-    # Kill all processes in the control group on stop/restart. This will kill
-    # all the jobs running, so ensure that we only upgrade when the nodes are
-    # not in use. See:
-    # https://github.com/NixOS/nixpkgs/commit/ae93ed0f0d4e7be0a286d1fca86446318c0c6ffb
-    # https://bugs.schedmd.com/show_bug.cgi?id=2095#c24
-    KillMode = lib.mkForce "control-group";
-  };
-  services.slurm = {
-    client.enable = true;
-    controlMachine = "hut";
-    clusterName = "jungle";
-    nodeName = [
-      "owl[1,2]  Sockets=2 CoresPerSocket=14 ThreadsPerCore=2 Feature=owl"
-      "hut       Sockets=2 CoresPerSocket=14 ThreadsPerCore=2"
-    ];
-
-    # See slurm.conf(5) for more details about these options.
-    extraConfig = ''
-      # Use PMIx for MPI by default. It works okay with MPICH and OpenMPI, but
-      # not with Intel MPI. For that use the compatibility shim libpmi.so
-      # setting I_MPI_PMI_LIBRARY=$pmix/lib/libpmi.so while maintaining the PMIx
-      # library in SLURM (--mpi=pmix). See more details here:
-      # https://pm.bsc.es/gitlab/rarias/jungle/-/issues/16
-      MpiDefault=pmix
-
-      # When a node reboots return that node to the slurm queue as soon as it
-      # becomes operative again.
-      ReturnToService=2
-
-      # Track all processes by using a cgroup
-      ProctrackType=proctrack/cgroup
-
-      # Enable task/affinity to allow the jobs to run in a specified subset of
-      # the resources. Use the task/cgroup plugin to enable process containment.
-      TaskPlugin=task/affinity,task/cgroup
-    '';
-  };
-}

m/common/ssh.nix

@@ -1,24 +0,0 @@
-{ ... }:
-
-{
-  # Enable the OpenSSH daemon.
-  services.openssh.enable = true;
-
-  # Connect to intranet git hosts via proxy
-  programs.ssh.extraConfig = ''
-    Host bscpm02.bsc.es bscpm03.bsc.es gitlab-internal.bsc.es alya.gitlab.bsc.es
-      User git
-      ProxyCommand nc -X connect -x localhost:23080 %h %p
-  '';
-
-  programs.ssh.knownHosts = {
-    "hut".publicKey = "ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAICO7jIp6JRnRWTMDsTB/aiaICJCl4x8qmKMPSs4lCqP1";
-    "owl1".publicKey = "ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIMqMEXO0ApVsBA6yjmb0xP2kWyoPDIWxBB0Q3+QbHVhv";
-    "owl2".publicKey = "ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIHurEYpQzNHqWYF6B9Pd7W8UPgF3BxEg0BvSbsA7BAdK";
-    "eudy".publicKey = "ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIL+WYPRRvZupqLAG0USKmd/juEPmisyyJaP8hAgYwXsG";
-    "koro".publicKey = "ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIImiTFDbxyUYPumvm8C4mEnHfuvtBY1H8undtd6oDd67";
-
-    "gitlab-internal.bsc.es".publicKey = "ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIF9arsAOSRB06hdy71oTvJHG2Mg8zfebADxpvc37lZo3";
-    "bscpm03.bsc.es".publicKey = "ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIM2NuSUPsEhqz1j5b4Gqd+MWFnRqyqY57+xMvBUqHYUS";
-  };
-}

m/common/xeon.nix

@@ -0,0 +1,9 @@
+{
+  # Provides the base system for a xeon node.
+  imports = [
+    ./base.nix
+    ./xeon/fs.nix
+    ./xeon/console.nix
+    ./xeon/net.nix
+  ];
+}

m/common/xeon/console.nix

@@ -0,0 +1,14 @@
+{
+  # Restart the serial console
+  systemd.services."serial-getty@ttyS0" = {
+    enable = true;
+    wantedBy = [ "getty.target" ];
+    serviceConfig.Restart = "always";
+  };
+
+  # Enable serial console
+  boot.kernelParams = [
+    "console=tty1"
+    "console=ttyS0,115200"
+  ];
+}

m/common/xeon/fs.nix

@@ -0,0 +1,8 @@
+{
+  # Mount the home via NFS
+  fileSystems."/home" = {
+    device = "10.0.40.30:/home";
+    fsType = "nfs";
+    options = [ "nfsvers=3" "rsize=1024" "wsize=1024" "cto" "nofail" ];
+  };
+}

m/common/xeon/net.nix

@@ -6,12 +6,11 @@
   boot.kernelModules = [ "ib_umad" "ib_ipoib" ];
 
   networking = {
-    enableIPv6 = false;
-    useDHCP = false;
     defaultGateway = "10.0.40.30";
     nameservers = ["8.8.8.8"];
+
     proxy = {
-      default = "http://localhost:23080/";
+      default = "http://hut:23080/";
       noProxy = "127.0.0.1,localhost,internal.domain,10.0.40.40";
       # Don't set all_proxy as go complains and breaks the gitlab runner, see:
       # https://github.com/golang/go/issues/16715
@@ -19,21 +18,23 @@
     };
 
     firewall = {
-      enable = true;
-      allowedTCPPorts = [ 22 ];
-
-      # FIXME: For slurmd as it requests the compute nodes to connect to us
-      allowedTCPPortRanges = [ { from=1024; to=65535; } ];
+      extraCommands = ''
+        # Prevent ssfhead from contacting our slurmd daemon
+        iptables -A nixos-fw -p tcp -s ssfhead --dport 6817:6819 -j nixos-fw-refuse
+        # But accept traffic to slurm ports from any other node in the subnet
+        iptables -A nixos-fw -p tcp -s 10.0.40.0/24 --dport 6817:6819 -j nixos-fw-accept
+        # We also need to open the srun port range
+        iptables -A nixos-fw -p tcp -s 10.0.40.0/24 --dport 60000:61000 -j nixos-fw-accept
+      '';
     };
 
     extraHosts = ''
-      10.0.40.30      ssfhead
-      84.88.53.236    ssfhead.bsc.es ssfhead
+      10.0.40.30              ssfhead
       
       # Node Entry for node: mds01 (ID=72)
       10.0.40.40              bay mds01 mds01-eth0
-      10.0.42.40              mds01-ib0
-      10.0.40.141             mds01-ipmi0
+      10.0.42.40              bay-ib mds01-ib0
+      10.0.40.141             bay-ipmi mds01-ipmi0
       
       # Node Entry for node: oss01 (ID=73)
       10.0.40.41              oss01 oss01-eth0
@@ -42,18 +43,18 @@
       
       # Node Entry for node: oss02 (ID=74)
       10.0.40.42              lake2 oss02 oss02-eth0
-      10.0.42.42              oss02-ib0
-      10.0.40.143             oss02-ipmi0
+      10.0.42.42              lake2-ib oss02-ib0
+      10.0.40.143             lake2-ipmi oss02-ipmi0
       
       # Node Entry for node: xeon01 (ID=15)
       10.0.40.1               owl1 xeon01 xeon01-eth0
-      10.0.42.1               xeon01-ib0
-      10.0.40.101             xeon01-ipmi0
+      10.0.42.1               owl1-ib xeon01-ib0
+      10.0.40.101             owl1-ipmi xeon01-ipmi0
       
       # Node Entry for node: xeon02 (ID=16)
       10.0.40.2               owl2 xeon02 xeon02-eth0
-      10.0.42.2               xeon02-ib0
-      10.0.40.102             xeon02-ipmi0
+      10.0.42.2               owl2-ib xeon02-ib0
+      10.0.40.102             owl2-ipmi xeon02-ipmi0
       
       # Node Entry for node: xeon03 (ID=17)
       10.0.40.3               xeon03 xeon03-eth0
@@ -67,8 +68,8 @@
       
       # Node Entry for node: xeon05 (ID=19)
       10.0.40.5               koro xeon05 xeon05-eth0
-      10.0.42.5               xeon05-ib0
-      10.0.40.105             xeon05-ipmi0
+      10.0.42.5               koro-ib xeon05-ib0
+      10.0.40.105             koro-ipmi xeon05-ipmi0
       
       # Node Entry for node: xeon06 (ID=20)
       10.0.40.6               xeon06 xeon06-eth0
@@ -77,13 +78,13 @@
       
       # Node Entry for node: xeon07 (ID=21)
       10.0.40.7               hut xeon07 xeon07-eth0
-      10.0.42.7               xeon07-ib0
-      10.0.40.107             xeon07-ipmi0
+      10.0.42.7               hut-ib xeon07-ib0
+      10.0.40.107             hut-ipmi xeon07-ipmi0
       
       # Node Entry for node: xeon08 (ID=22)
       10.0.40.8               eudy xeon08 xeon08-eth0
-      10.0.42.8               xeon08-ib0
-      10.0.40.108             xeon08-ipmi0
+      10.0.42.8               eudy-ib xeon08-ib0
+      10.0.40.108             eudy-ipmi xeon08-ipmi0
     '';
   };
 }

m/eudy/configuration.nix

@@ -2,14 +2,14 @@
 
 {
   imports = [
-    ../common/main.nix
+    ../common/xeon.nix
     #(modulesPath + "/installer/netboot/netboot-minimal.nix")
 
     ./kernel/kernel.nix
     ./cpufreq.nix
     ./fs.nix
     ./users.nix
-    ./slurm.nix
+    ../module/debuginfod.nix
   ];
 
   # Select this using the ID to avoid mismatches

m/eudy/kernel/kernel.nix

@@ -21,9 +21,9 @@ let
   #   configfile = if lockdep then ./configs/lockdep else ./configs/defconfig;
   #};
 
-  kernel = nixos-fcsv3;
+  kernel = nixos-fcs;
 
-  nixos-fcs-kernel = {gitCommit, lockStat ? false, preempt ? false, branch ? "fcs"}: pkgs.linuxPackagesFor (pkgs.buildLinux rec {
+  nixos-fcs-kernel = lib.makeOverridable ({gitCommit, lockStat ? false, preempt ? false, branch ? "fcs"}: pkgs.linuxPackagesFor (pkgs.buildLinux rec {
     version = "6.2.8";
     src = builtins.fetchGit {
       url = "git@bscpm03.bsc.es:ompss-kernel/linux.git";
@@ -40,35 +40,13 @@ let
     };
     kernelPatches = [];
     extraMeta.branch = lib.versions.majorMinor version;
-  });
+  }));
 
-  nixos-fcsv1 = nixos-fcs-kernel {gitCommit = "bc11660676d3d68ce2459b9fb5d5e654e3f413be";};
-  nixos-fcsv2 = nixos-fcs-kernel {gitCommit = "db0f2eca0cd57a58bf456d7d2c7d5d8fdb25dfb1";};
-  nixos-fcsv3 = nixos-fcs-kernel {gitCommit = "6c17394890704c3345ac1a521bb547164b36b154";};
-
-  # always use fcs_sched_setaffinity
-  #nixos-debug = nixos-fcs-kernel {gitCommit = "7d0bf285fca92badc8df3c9907a9ab30db4418aa";};
-  # remove need_check_cgroup
-  #nixos-debug = nixos-fcs-kernel {gitCommit = "4cc4efaab5e4a0bfa3089e935215b981c1922919";};
-  # merge again fcs_wake and fcs_wait
-  #nixos-debug = nixos-fcs-kernel {gitCommit = "40c6f72f4ae54b0b636b193ac0648fb5730c810d";};
-  # start from scratch, this is the working version with split fcs_wake and fcs_wait
-  nixos-debug = nixos-fcs-kernel {gitCommit = "c9a39d6a4ca83845b4e71fcc268fb0a76aff1bdf"; branch = "fcs-test"; };
-
-  nixos-fcsv1-lockstat = nixos-fcs-kernel {
-    gitCommit = "bc11660676d3d68ce2459b9fb5d5e654e3f413be";
+  nixos-fcs = nixos-fcs-kernel {gitCommit = "8a09822dfcc8f0626b209d6d2aec8b5da459dfee";};
+  nixos-fcs-lockstat = nixos-fcs.override {
     lockStat = true;
   };
-  nixos-fcsv2-lockstat = nixos-fcs-kernel {
-    gitCommit = "db0f2eca0cd57a58bf456d7d2c7d5d8fdb25dfb1";
-    lockStat = true;
-  };
-  nixos-fcsv3-lockstat = nixos-fcs-kernel {
-    gitCommit = "6c17394890704c3345ac1a521bb547164b36b154";
-    lockStat = true;
-  };
-  nixos-fcsv3-lockstat-preempt = nixos-fcs-kernel {
-    gitCommit = "6c17394890704c3345ac1a521bb547164b36b154";
+  nixos-fcs-lockstat-preempt = nixos-fcs.override {
     lockStat = true;
     preempt = true;
   };

m/eudy/slurm.nix

@@ -1,7 +0,0 @@
-{ lib, ... }:
-
-{
-  services.slurm = {
-    client.enable = lib.mkForce false;
-  };
-}

m/hut/blackbox.yml

@@ -0,0 +1,162 @@
+modules:
+  http_2xx:
+    prober: http
+    timeout: 5s
+    http:
+      proxy_url: "http://127.0.0.1:23080"
+      skip_resolve_phase_with_proxy: true
+      follow_redirects: true
+      valid_status_codes: []  # Defaults to 2xx
+      method: GET
+  http_with_proxy:
+    prober: http
+    http:
+      proxy_url: "http://127.0.0.1:3128"
+      skip_resolve_phase_with_proxy: true
+  http_with_proxy_and_headers:
+    prober: http
+    http:
+      proxy_url: "http://127.0.0.1:3128"
+      proxy_connect_header:
+        Proxy-Authorization:
+          - Bearer token
+  http_post_2xx:
+    prober: http
+    timeout: 5s
+    http:
+      method: POST
+      headers:
+        Content-Type: application/json
+      body: '{}'
+  http_post_body_file:
+    prober: http
+    timeout: 5s
+    http:
+      method: POST
+      body_file: "/files/body.txt"
+  http_basic_auth_example:
+    prober: http
+    timeout: 5s
+    http:
+      method: POST
+      headers:
+        Host: "login.example.com"
+      basic_auth:
+        username: "username"
+        password: "mysecret"
+  http_2xx_oauth_client_credentials:
+    prober: http
+    timeout: 5s
+    http:
+      valid_http_versions: ["HTTP/1.1", "HTTP/2"]
+      follow_redirects: true
+      preferred_ip_protocol: "ip4"
+      valid_status_codes:
+        - 200
+        - 201
+      oauth2:
+        client_id: "client_id"
+        client_secret: "client_secret"
+        token_url: "https://api.example.com/token"
+        endpoint_params:
+          grant_type: "client_credentials"
+  http_custom_ca_example:
+    prober: http
+    http:
+      method: GET
+      tls_config:
+        ca_file: "/certs/my_cert.crt"
+  http_gzip:
+    prober: http
+    http:
+      method: GET
+      compression: gzip
+  http_gzip_with_accept_encoding:
+    prober: http
+    http:
+      method: GET
+      compression: gzip
+      headers:
+        Accept-Encoding: gzip
+  tls_connect:
+    prober: tcp
+    timeout: 5s
+    tcp:
+      tls: true
+  tcp_connect_example:
+    prober: tcp
+    timeout: 5s
+  imap_starttls:
+    prober: tcp
+    timeout: 5s
+    tcp:
+      query_response:
+        - expect: "OK.*STARTTLS"
+        - send: ". STARTTLS"
+        - expect: "OK"
+        - starttls: true
+        - send: ". capability"
+        - expect: "CAPABILITY IMAP4rev1"
+  smtp_starttls:
+    prober: tcp
+    timeout: 5s
+    tcp:
+      query_response:
+        - expect: "^220 ([^ ]+) ESMTP (.+)$"
+        - send: "EHLO prober\r"
+        - expect: "^250-STARTTLS"
+        - send: "STARTTLS\r"
+        - expect: "^220"
+        - starttls: true
+        - send: "EHLO prober\r"
+        - expect: "^250-AUTH"
+        - send: "QUIT\r"
+  irc_banner_example:
+    prober: tcp
+    timeout: 5s
+    tcp:
+      query_response:
+        - send: "NICK prober"
+        - send: "USER prober prober prober :prober"
+        - expect: "PING :([^ ]+)"
+          send: "PONG ${1}"
+        - expect: "^:[^ ]+ 001"
+  icmp:
+    prober: icmp
+    timeout: 5s
+    icmp:
+      preferred_ip_protocol: "ip4"
+  dns_udp_example:
+    prober: dns
+    timeout: 5s
+    dns:
+      query_name: "www.prometheus.io"
+      query_type: "A"
+      valid_rcodes:
+        - NOERROR
+      validate_answer_rrs:
+        fail_if_matches_regexp:
+          - ".*127.0.0.1"
+        fail_if_all_match_regexp:
+          - ".*127.0.0.1"
+        fail_if_not_matches_regexp:
+          - "www.prometheus.io.\t300\tIN\tA\t127.0.0.1"
+        fail_if_none_matches_regexp:
+          - "127.0.0.1"
+      validate_authority_rrs:
+        fail_if_matches_regexp:
+          - ".*127.0.0.1"
+      validate_additional_rrs:
+        fail_if_matches_regexp:
+          - ".*127.0.0.1"
+  dns_soa:
+    prober: dns
+    dns:
+      query_name: "prometheus.io"
+      query_type: "SOA"
+  dns_tcp_example:
+    prober: dns
+    dns:
+      transport_protocol: "tcp" # defaults to "udp"
+      preferred_ip_protocol: "ip4" # defaults to "ip6"
+      query_name: "www.prometheus.io"

m/hut/configuration.nix

@@ -1,20 +1,23 @@
-{ config, pkgs, agenix, ... }:
+{ config, pkgs, ... }:
 
 {
   imports = [
-    ../common/main.nix
+    ../common/xeon.nix
 
+    ../module/ceph.nix
+    ../module/debuginfod.nix
+    ../module/emulation.nix
+    ../module/slurm-client.nix
     ./gitlab-runner.nix
     ./monitoring.nix
     ./nfs.nix
-    ./slurm-daemon.nix
-    ./ceph.nix
+    ./slurm-server.nix
+    ./nix-serve.nix
+    ./public-inbox.nix
+    ./gitea.nix
+    ./msmtp.nix
+    ./postgresql.nix
     #./pxe.nix
-    agenix.nixosModules.default
-  ];
-
-  environment.systemPackages = [
-    agenix.packages.x86_64-linux.default
   ];
 
   # Select the this using the ID to avoid mismatches
@@ -30,5 +33,15 @@
       address = "10.0.42.7";
       prefixLength = 24;
     } ];
+    firewall = {
+      extraCommands = ''
+        # Accept all proxy traffic from compute nodes but not the login
+        iptables -A nixos-fw -p tcp -s 10.0.40.30 --dport 23080 -j nixos-fw-log-refuse
+        iptables -A nixos-fw -p tcp -s 10.0.40.0/24 --dport 23080 -j nixos-fw-accept
+      '';
+    };
   };
+
+  # Allow proxy to bind to the ethernet interface
+  services.openssh.settings.GatewayPorts = "clientspecified";
 }

m/hut/gitea.nix

@@ -0,0 +1,63 @@
+{ config, lib, ... }:
+{
+  age.secrets.giteaRunnerToken.file = ../../secrets/gitea-runner-token.age;
+
+  services.gitea = {
+    enable = true;
+    appName = "Gitea in the jungle";
+
+    settings = {
+      server = {
+        ROOT_URL = "https://jungle.bsc.es/git/";
+        LOCAL_ROOT_URL = "https://jungle.bsc.es/git/";
+        LANDING_PAGE = "explore";
+      };
+      metrics.ENABLED = true;
+      service = {
+        REGISTER_MANUAL_CONFIRM = true;
+        ENABLE_NOTIFY_MAIL = true;
+      };
+      log.LEVEL = "Warn";
+
+      mailer = {
+        ENABLED       = true;
+        FROM          = "jungle-robot@bsc.es";
+        PROTOCOL      = "sendmail";
+        SENDMAIL_PATH = "/run/wrappers/bin/sendmail";
+        SENDMAIL_ARGS = "--";
+      };
+    };
+  };
+
+  services.gitea-actions-runner.instances = {
+    runrun = {
+      enable = true;
+      name = "runrun";
+      url = "https://jungle.bsc.es/git/";
+      tokenFile = config.age.secrets.giteaRunnerToken.path;
+      labels = [ "native:host" ];
+      settings.runner.capacity = 8;
+    };
+  };
+
+  systemd.services.gitea-runner-runrun = {
+    path = [ "/run/current-system/sw" ];
+    serviceConfig = {
+      # DynamicUser doesn't work well with SSH
+      DynamicUser = lib.mkForce false;
+      User = "gitea-runner";
+      Group = "gitea-runner";
+    };
+  };
+
+  users.users.gitea-runner = {
+    isSystemUser = true;
+    home = "/var/lib/gitea-runner";
+    description = "Gitea Runner";
+    group = "gitea-runner";
+    extraGroups = [ "docker" ];
+    createHome = true;
+  };
+  users.groups.gitea-runner = {};
+}
+

m/hut/gitlab-runner.nix

@@ -1,44 +1,38 @@
 { pkgs, lib, config, ... }:
 
 {
-  age.secrets."secrets/ovni-token".file = ./secrets/ovni-token.age;
-  age.secrets."secrets/nosv-token".file = ./secrets/nosv-token.age;
+  age.secrets.gitlabRunnerShellToken.file = ../../secrets/gitlab-runner-shell-token.age;
+  age.secrets.gitlabRunnerDockerToken.file = ../../secrets/gitlab-runner-docker-token.age;
 
   services.gitlab-runner = {
     enable = true;
     settings.concurrent = 5;
-    services = {
-      ovni-shell = {
-        registrationConfigFile = config.age.secrets."secrets/ovni-token".path;
+    services = let
+      common-shell = {
         executor = "shell";
-        tagList = [ "nix" "xeon" ];
         environmentVariables = {
           SHELL = "${pkgs.bash}/bin/bash";
         };
       };
-      ovni-docker = {
-        registrationConfigFile = config.age.secrets."secrets/ovni-token".path;
+      common-docker = {
+        executor = "docker";
         dockerImage = "debian:stable";
-        tagList = [ "docker" "xeon" ];
-        registrationFlags = [ "--docker-network-mode host" ];
-        environmentVariables = {
-          https_proxy = "http://localhost:23080";
-          http_proxy = "http://localhost:23080";
-        };
-      };
-      nosv-docker = {
-        registrationConfigFile = config.age.secrets."secrets/nosv-token".path;
-        dockerImage = "debian:stable";
-        tagList = [ "docker" "xeon" ];
         registrationFlags = [
           "--docker-network-mode host"
-          "--docker-cpus 56"
         ];
         environmentVariables = {
           https_proxy = "http://localhost:23080";
           http_proxy = "http://localhost:23080";
         };
       };
+    in {
+      # For pm.bsc.es/gitlab
+      gitlab-pm-shell = common-shell // {
+        authenticationTokenConfigFile = config.age.secrets.gitlabRunnerShellToken.path;
+      };
+      gitlab-pm-docker = common-docker // {
+        authenticationTokenConfigFile = config.age.secrets.gitlabRunnerDockerToken.path;
+      };
     };
   };
 

m/hut/monitoring.nix

@@ -1,6 +1,14 @@
 { config, lib, ... }:
 
 {
+  imports = [ ../module/slurm-exporter.nix ];
+
+  age.secrets.grafanaJungleRobotPassword = {
+    file = ../../secrets/jungle-robot-password.age;
+    owner = "grafana";
+    mode = "400";
+  };
+
   services.grafana = {
     enable = true;
     settings = {
@@ -11,14 +19,30 @@
         http_port = 2342;
         http_addr = "127.0.0.1";
       };
+      smtp = {
+        enabled = true;
+        from_address = "jungle-robot@bsc.es";
+        user = "jungle-robot";
+        # Read the password from a file, which is only readable by grafana user
+        # https://grafana.com/docs/grafana/latest/setup-grafana/configure-grafana/#file-provider
+        password = "$__file{${config.age.secrets.grafanaJungleRobotPassword.path}}";
+        host = "mail.bsc.es:465";
+        startTLS_policy = "NoStartTLS";
+      };
       feature_toggles.publicDashboards = true;
+      "auth.anonymous".enabled = true;
+      log.level = "warn";
     };
   };
 
+  # Make grafana alerts also use the proxy
+  systemd.services.grafana.environment = config.networking.proxy.envVars;
+
   services.prometheus = {
     enable = true;
     port = 9001;
     retentionTime = "1y";
+    listenAddress = "127.0.0.1";
   };
 
   systemd.services.prometheus-ipmi-exporter.serviceConfig.DynamicUser = lib.mkForce false;
@@ -48,13 +72,23 @@
         user = "root";
         configFile = ./ipmi.yml;
         #extraFlags = [ "--log.level=debug" ];
+        listenAddress = "127.0.0.1";
       };
       node = {
         enable = true;
         enabledCollectors = [ "systemd" ];
         port = 9002;
+        listenAddress = "127.0.0.1";
+      };
+      smartctl = {
+        enable = true;
+        listenAddress = "127.0.0.1";
+      };
+      blackbox = {
+        enable = true;
+        listenAddress = "127.0.0.1";
+        configFile = ./blackbox.yml;
       };
-      smartctl.enable = true;
     };
 
     scrapeConfigs = [
@@ -67,6 +101,8 @@
             "127.0.0.1:9323"
             "127.0.0.1:9252"
             "127.0.0.1:${toString config.services.prometheus.exporters.smartctl.port}"
+            "127.0.0.1:9341" # Slurm exporter
+            "127.0.0.1:${toString config.services.prometheus.exporters.blackbox.port}"
           ];
         }];
       }
@@ -80,6 +116,71 @@
           ];
         }];
       }
+      {
+        job_name = "blackbox-http";
+        metrics_path = "/probe";
+        params = { module = [ "http_2xx" ]; };
+        static_configs = [{
+          targets = [
+            "https://www.google.com/robots.txt"
+            "https://pm.bsc.es/"
+            "https://pm.bsc.es/gitlab/"
+            "https://jungle.bsc.es/"
+            "https://gitlab.bsc.es/"
+          ];
+        }];
+        relabel_configs = [
+          {
+            # Takes the address and sets it in the "target=<xyz>" URL parameter
+            source_labels = [ "__address__" ];
+            target_label = "__param_target";
+          }
+          {
+            # Sets the "instance" label with the remote host we are querying
+            source_labels = [ "__param_target" ];
+            target_label = "instance";
+          }
+          {
+            # Shows the host target address instead of the blackbox address
+            target_label = "__address__";
+            replacement = "127.0.0.1:${toString config.services.prometheus.exporters.blackbox.port}";
+          }
+        ];
+      }
+      {
+        job_name = "blackbox-icmp";
+        metrics_path = "/probe";
+        params = { module = [ "icmp" ]; };
+        static_configs = [{
+          targets = [
+            "1.1.1.1"
+            "8.8.8.8"
+            "ssfhead"
+            "anella-bsc.cesca.cat"
+          ];
+        }];
+        relabel_configs = [
+          {
+            # Takes the address and sets it in the "target=<xyz>" URL parameter
+            source_labels = [ "__address__" ];
+            target_label = "__param_target";
+          }
+          {
+            # Sets the "instance" label with the remote host we are querying
+            source_labels = [ "__param_target" ];
+            target_label = "instance";
+          }
+          {
+            # Shows the host target address instead of the blackbox address
+            target_label = "__address__";
+            replacement = "127.0.0.1:${toString config.services.prometheus.exporters.blackbox.port}";
+          }
+        ];
+      }
+      {
+        job_name = "gitea";
+        static_configs = [{ targets = [ "127.0.0.1:3000" ]; }];
+      }
       {
         # Scrape the IPMI info of the hosts remotely via LAN
         job_name = "ipmi-lan";
@@ -132,6 +233,17 @@
           }
         ];
       }
+      {
+        job_name = "ipmi-raccoon";
+        metrics_path = "/ipmi";
+        static_configs = [
+          { targets = [ "127.0.0.1:9291" ]; }
+        ];
+        params = {
+          target = [ "84.88.51.142" ];
+          module = [ "raccoon" ];
+        };
+      }
     ];
   };
 }

m/hut/msmtp.nix

@@ -0,0 +1,24 @@
+{ config, lib, ... }:
+{
+  age.secrets.jungleRobotPassword = {
+    file = ../../secrets/jungle-robot-password.age;
+    group = "gitea";
+    mode = "440";
+  };
+
+  programs.msmtp = {
+    enable = true;
+    accounts = {
+      default = {
+        auth = true;
+        tls = true;
+        tls_starttls = false;
+        port = 465;
+        host = "mail.bsc.es";
+        user = "jungle-robot";
+        passwordeval = "cat ${config.age.secrets.jungleRobotPassword.path}";
+        from = "jungle-robot@bsc.es";
+      };
+    };
+  };
+}

m/hut/nix-serve.nix

@@ -0,0 +1,16 @@
+{ config, ... }:
+
+{
+  age.secrets.nixServe.file = ../../secrets/nix-serve.age;
+
+  services.nix-serve = {
+    enable = true;
+    # Only listen locally, as we serve it via ssh
+    bindAddress = "127.0.0.1";
+    port = 5000;
+
+    secretKeyFile = config.age.secrets.nixServe.path;
+    # Public key:
+    # jungle.bsc.es:pEc7MlAT0HEwLQYPtpkPLwRsGf80ZI26aj29zMw/HH0=
+  };
+}

m/hut/postgresql.nix

@@ -0,0 +1,19 @@
+{ lib, ... }:
+
+{
+  services.postgresql = {
+    enable = true;
+    ensureDatabases = [ "perftestsdb" ];
+    ensureUsers = [
+      { name = "anavarro"; ensureClauses.superuser = true; }
+      { name = "rarias";   ensureClauses.superuser = true; }
+      { name = "grafana"; }
+    ];
+    authentication = ''
+      #type  database     DBuser    auth-method
+      local  perftestsdb  rarias    trust
+      local  perftestsdb  anavarro  trust
+      local  perftestsdb  grafana   trust
+    '';
+  };
+}

m/hut/public-inbox.css

@@ -0,0 +1,79 @@
+/*
+ * CC0-1.0 <https://creativecommons.org/publicdomain/zero/1.0/legalcode>
+ * Dark color scheme using 216 web-safe colors, inspired
+ * somewhat by the default color scheme in mutt.
+ * It reduces eyestrain for me, and energy usage for all:
+ * https://en.wikipedia.org/wiki/Light-on-dark_color_scheme
+ */
+
+* {
+	font-size: 14px;
+	font-family: monospace;
+}
+
+pre {
+	white-space: pre-wrap;
+	padding: 10px;
+	background: #f5f5f5;
+}
+
+hr {
+	margin: 30px 0;
+}
+
+body {
+	max-width: 120ex; /* 120 columns wide */
+	margin: 50px auto;
+}
+
+/*
+ * Underlined links add visual noise which make them hard-to-read.
+ * Use colors to make them stand out, instead.
+ */
+a:link {
+	color: #007;
+	text-decoration: none;
+}
+a:visited {
+	color:#504;
+}
+a:hover {
+	text-decoration: underline;
+}
+
+/* quoted text in emails gets a different color */
+*.q { color:gray }
+
+/*
+ * these may be used with cgit <https://git.zx2c4.com/cgit/>, too.
+ * (cgit uses <div>, public-inbox uses <span>)
+ */
+*.add { color:darkgreen } /* diff post-image lines */
+*.del { color:darkred } /* diff pre-image lines */
+*.head { color:black } /* diff header (metainformation) */
+*.hunk { color:gray } /* diff hunk-header */
+
+/*
+ * highlight 3.x colors (tested 3.18) for displaying blobs.
+ * This doesn't use most of the colors available, as I find too
+ * many colors overwhelming, so the default is commented out.
+ */
+.hl.num { color:#f30 } /* number */
+.hl.esc { color:#f0f } /* escape character */
+.hl.str { color:#f30 } /* string */
+.hl.ppc { color:#f0f } /* preprocessor */
+.hl.pps { color:#f30 } /* preprocessor string */
+.hl.slc { color:#09f } /* single-line comment */
+.hl.com { color:#09f } /* multi-line comment */
+/* .hl.opt { color:#ccc } */ /* operator */
+/* .hl.ipl { color:#ccc } */ /* interpolation */
+
+/* keyword groups kw[a-z] */
+.hl.kwa { color:#ff0 }
+.hl.kwb { color:#0f0 }
+.hl.kwc { color:#ff0 }
+/* .hl.kwd { color:#ccc } */
+
+/* line-number (unused by public-inbox) */
+/* .hl.lin { color:#ccc } */
+

m/hut/public-inbox.nix

@@ -0,0 +1,47 @@
+{ lib, ... }:
+
+{
+  services.public-inbox = {
+    enable = true;
+    http = {
+      enable = true;
+      port = 8081;
+      mounts = [ "/lists" ];
+    };
+    settings.publicinbox = {
+      css = [ "${./public-inbox.css}" ];
+      wwwlisting = "all";
+    };
+    inboxes = {
+      bscpkgs = {
+        url = "https://jungle.bsc.es/lists/bscpkgs";
+        address = [ "~rodarima/bscpkgs@lists.sr.ht" ];
+        watch = [ "imaps://jungle-robot%40gmx.com@imap.gmx.com/INBOX" ];
+        description = "Patches for bscpkgs";
+        listid = "~rodarima/bscpkgs.lists.sr.ht";
+      };
+      jungle = {
+        url = "https://jungle.bsc.es/lists/jungle";
+        address = [ "~rodarima/jungle@lists.sr.ht" ];
+        watch = [ "imaps://jungle-robot%40gmx.com@imap.gmx.com/INBOX" ];
+        description = "Patches for jungle";
+        listid = "~rodarima/jungle.lists.sr.ht";
+      };
+    };
+  };
+
+  # We need access to the network for the watch service, as we will fetch the
+  # emails directly from the IMAP server.
+  systemd.services.public-inbox-watch.serviceConfig = {
+    PrivateNetwork = lib.mkForce false;
+    RestrictAddressFamilies = lib.mkForce [ "AF_UNIX"  "AF_INET" "AF_INET6" ];
+    KillSignal = "SIGKILL"; # Avoid slow shutdown
+
+    # Required for chmod(..., 02750) on directories by git, from
+    # systemd.exec(8):
+    # > Note that this restricts marking of any type of file system object with
+    # > these bits, including both regular files and directories (where the SGID
+    # > is a different meaning than for files, see documentation).
+    RestrictSUIDSGID = lib.mkForce false;
+  };
+}

m/hut/secrets.nix

@@ -1,11 +0,0 @@
-let
-  rarias  = "ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIE1oZTPtlEXdGt0Ak+upeCIiBdaDQtcmuWoTUCVuSVIR rarias@hut";
-  root    = "ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIII/1TNArcwA6D47mgW4TArwlxQRpwmIGiZDysah40Gb";
-  hut     = "ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAICO7jIp6JRnRWTMDsTB/aiaICJCl4x8qmKMPSs4lCqP1";
-  default = [ rarias root hut ];
-in
-{
-  "secrets/ovni-token.age".publicKeys = default;
-  "secrets/nosv-token.age".publicKeys = default;
-  "secrets/ceph-user.age".publicKeys = default;
-}

m/hut/secrets/ceph-user.age

@@ -1,11 +0,0 @@
-age-encryption.org/v1
--> ssh-ed25519 CAWG4Q 35Ak+Mep9k5KnDLF1ywDbMD4l4mRFg6D0et19tqXxAw
-Wgr+CX4rzrPmUszSidtLAVSvgD80F2dqtd92hGZIFwo
--> ssh-ed25519 MSF3dg OVFvpkAyWTowtxsafstX31H/hJpNZmnOCbvqMIN0+AQ
-VxjRcQmp+BadEh2y0PB96EeizIl3tTQpVu0CWHmsc1s
--> ssh-ed25519 HY2yRg MJSQIpre9m0XnojgXuKQ/+hVBZNrZNGZqplwhqicpjI
-CLkE52iqpoqSnbzisNjQgxTfNqKeaRl5ntcw1d+ZDyQ
--> m$8`De%~-grease '85p}`by
-52zMpprONcawWDDtzHdWNwFoYXErPUnVjhSONbUBpDlqAmJmD1LcAnsU
---- 0vZOPyXQIMMGTwgFfvm8Sn8O7vjrsjGUEy5m/BASCyc
-È| üœ)‡<>ËëË*_ËDóUS`<06><>‹àŠèr Âs<C382>¢NªÈ[ÖŒ^e+A1œ“G.í#âù°m˜¸Wß ’5·àƒµ( 

m/hut/slurm-daemon.nix

@@ -1,11 +0,0 @@
-{ ... }:
-
-{
-  services.slurm = {
-    server.enable = true;
-    partitionName = [
-      "owl Nodes=owl[1-2] Default=YES MaxTime=INFINITE State=UP"
-      "all Nodes=owl[1-2],hut Default=NO MaxTime=INFINITE State=UP"
-    ];
-  };
-}

m/hut/slurm-server.nix

@@ -0,0 +1,7 @@
+{ ... }:
+
+{
+  services.slurm = {
+    server.enable = true;
+  };
+}

m/hut/targets.yml

@@ -7,5 +7,9 @@
   - 10.0.40.106
   - 10.0.40.107
   - 10.0.40.108
+  # Storage
+  - 10.0.40.141
+  - 10.0.40.142
+  - 10.0.40.143
   labels:
     job: ipmi-lan

m/koro/configuration.nix

@@ -2,13 +2,11 @@
 
 {
   imports = [
-    ../common/main.nix
+    ../common/xeon.nix
     #(modulesPath + "/installer/netboot/netboot-minimal.nix")
 
     ../eudy/cpufreq.nix
     ../eudy/users.nix
-    ../eudy/slurm.nix
-    ./users.nix
     ./kernel.nix
   ];
 

m/koro/kernel.nix

@@ -1,9 +1,29 @@
 { pkgs, lib, ... }:
 
 let
-  kernel = nixos-fcsv4;
+  #fcs-devel = pkgs.linuxPackages_custom {
+  #   version = "6.2.8";
+  #   src = /mnt/data/kernel/fcs/kernel/src;
+  #   configfile = /mnt/data/kernel/fcs/kernel/configs/defconfig;
+  #};
 
-  nixos-fcs-kernel = {gitCommit, lockStat ? false, preempt ? false, branch ? "fcs"}: pkgs.linuxPackagesFor (pkgs.buildLinux rec {
+  #fcsv1 = fcs-kernel "bc11660676d3d68ce2459b9fb5d5e654e3f413be" false;
+  #fcsv2 = fcs-kernel "db0f2eca0cd57a58bf456d7d2c7d5d8fdb25dfb1" false;
+  #fcsv1-lockdep = fcs-kernel "bc11660676d3d68ce2459b9fb5d5e654e3f413be" true;
+  #fcsv2-lockdep = fcs-kernel "db0f2eca0cd57a58bf456d7d2c7d5d8fdb25dfb1" true;
+  #fcs-kernel = gitCommit: lockdep: pkgs.linuxPackages_custom {
+  #   version = "6.2.8";
+  #   src = builtins.fetchGit {
+  #     url = "git@bscpm03.bsc.es:ompss-kernel/linux.git";
+  #     rev = gitCommit;
+  #     ref = "fcs";
+  #   };
+  #   configfile = if lockdep then ./configs/lockdep else ./configs/defconfig;
+  #};
+
+  kernel = nixos-fcs;
+
+  nixos-fcs-kernel = lib.makeOverridable ({gitCommit, lockStat ? false, preempt ? false, branch ? "fcs"}: pkgs.linuxPackagesFor (pkgs.buildLinux rec {
     version = "6.2.8";
     src = builtins.fetchGit {
       url = "git@bscpm03.bsc.es:ompss-kernel/linux.git";
@@ -20,27 +40,13 @@ let
     };
     kernelPatches = [];
     extraMeta.branch = lib.versions.majorMinor version;
-  });
+  }));
 
-  nixos-fcsv1 = nixos-fcs-kernel {gitCommit = "bc11660676d3d68ce2459b9fb5d5e654e3f413be";};
-  nixos-fcsv2 = nixos-fcs-kernel {gitCommit = "db0f2eca0cd57a58bf456d7d2c7d5d8fdb25dfb1";};
-  nixos-fcsv3 = nixos-fcs-kernel {gitCommit = "6c17394890704c3345ac1a521bb547164b36b154";};
-  nixos-fcsv4 = nixos-fcs-kernel {gitCommit = "c94c3d946f33ac3e5782a02ee002cc1164c0cb4f";};
-
-  nixos-fcsv1-lockstat = nixos-fcs-kernel {
-    gitCommit = "bc11660676d3d68ce2459b9fb5d5e654e3f413be";
+  nixos-fcs = nixos-fcs-kernel {gitCommit = "8a09822dfcc8f0626b209d6d2aec8b5da459dfee";};
+  nixos-fcs-lockstat = nixos-fcs.override {
     lockStat = true;
   };
-  nixos-fcsv2-lockstat = nixos-fcs-kernel {
-    gitCommit = "db0f2eca0cd57a58bf456d7d2c7d5d8fdb25dfb1";
-    lockStat = true;
-  };
-  nixos-fcsv3-lockstat = nixos-fcs-kernel {
-    gitCommit = "6c17394890704c3345ac1a521bb547164b36b154";
-    lockStat = true;
-  };
-  nixos-fcsv3-lockstat-preempt = nixos-fcs-kernel {
-    gitCommit = "6c17394890704c3345ac1a521bb547164b36b154";
+  nixos-fcs-lockstat-preempt = nixos-fcs.override {
     lockStat = true;
     preempt = true;
   };
@@ -60,5 +66,5 @@ in {
   
   # enable memory overcommit, needed to build a taglibc system using nix after
   # increasing the openblas memory footprint
-  boot.kernel.sysctl."vm.overcommit_memory" = lib.mkForce 1;
+  boot.kernel.sysctl."vm.overcommit_memory" = 1;
 }

m/koro/users.nix

@@ -1,17 +0,0 @@
-{ ... }:
-
-{
-  users.users = {
-    vlopez = {
-      uid = 4334;
-      isNormalUser = true;
-      home = "/home/Computational/vlopez";
-      description = "Victor López";
-      group = "Computational";
-      hashedPassword = "$6$0ZBkgIYE/renVqtt$1uWlJsb0FEezRVNoETTzZMx4X2SvWiOsKvi0ppWCRqI66S6TqMBXBdP4fcQyvRRBt0e4Z7opZIvvITBsEtO0f0";
-      openssh.authorizedKeys.keys = [
-        "ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIGMwlUZRf9jfG666Qa5Sb+KtEhXqkiMlBV2su3x/dXHq victor@arch"
-      ];
-    };
-  };
-}

m/lake2/configuration.nix

@@ -2,28 +2,27 @@
 
 {
   imports = [
-    ../common/main.nix
-    ../common/monitoring.nix
+    ../common/xeon.nix
+    ../module/monitoring.nix
   ];
 
   boot.loader.grub.device = "/dev/disk/by-id/wwn-0x55cd2e414d53563a";
 
+  boot.kernel.sysctl = {
+    "kernel.yama.ptrace_scope" = lib.mkForce "1";
+  };
+
   environment.systemPackages = with pkgs; [
     ceph
   ];
 
-  services.slurm = {
-    client.enable = lib.mkForce false;
-  };
-
   services.ceph = {
     enable = true;
     global = {
       fsid = "9c8d06e0-485f-4aaf-b16b-06d6daf1232b";
-      monHost = "10.0.42.40";
-      monInitialMembers = "10.0.42.40";
-      publicNetwork = "10.0.42.40/24";
-      clusterNetwork = "10.0.42.40/24";
+      monHost = "10.0.40.40";
+      monInitialMembers = "bay";
+      clusterNetwork = "10.0.40.40/24"; # Use Ethernet only
     };
     osd = {
       enable = true;
@@ -50,6 +49,16 @@
       address = "10.0.42.42";
       prefixLength = 24;
     } ];
+    firewall = {
+      extraCommands = ''
+        # Accept all incoming TCP traffic from bay
+        iptables -A nixos-fw -p tcp -s bay -j nixos-fw-accept
+        # Accept monitoring requests from hut
+        iptables -A nixos-fw -p tcp -s hut --dport 9002 -j nixos-fw-accept
+        # Accept all Ceph traffic from the local network
+        iptables -A nixos-fw -p tcp -s 10.0.40.0/24 -m multiport --dport 3300,6789,6800:7568 -j nixos-fw-accept
+      '';
+    };
   };
 
   # Missing service for volumes, see:

m/module/ceph.nix

@@ -1,8 +1,8 @@
 { config, pkgs, ... }:
 
+# Mounts the /ceph filesystem at boot
 {
   environment.systemPackages = with pkgs; [
-    ceph
     ceph-client
     fio # For benchmarks
   ];
@@ -11,14 +11,14 @@
   # modprobe command.
   boot.kernelModules = [ "ceph" ];
 
-  age.secrets."secrets/ceph-user".file = ./secrets/ceph-user.age;
+  age.secrets.cephUser.file = ../../secrets/ceph-user.age;
 
   fileSystems."/ceph" = {
     fsType = "ceph";
     device = "user@9c8d06e0-485f-4aaf-b16b-06d6daf1232b.cephfs=/";
     options = [
-      "mon_addr=10.0.42.40"
-      "secretfile=${config.age.secrets."secrets/ceph-user".path}"
+      "mon_addr=10.0.40.40"
+      "secretfile=${config.age.secrets.cephUser.path}"
     ];
   };
 }

m/module/debuginfod.nix

@@ -0,0 +1,3 @@
+{
+  services.nixseparatedebuginfod.enable = true;
+}

m/module/emulation.nix

@@ -0,0 +1,3 @@
+{
+  boot.binfmt.emulatedSystems = [ "armv7l-linux" "aarch64-linux" "powerpc64le-linux" "riscv64-linux" ];
+}

m/module/jungle-users.nix

@@ -0,0 +1,24 @@
+{ config, lib, ... }:
+
+with lib;
+
+{
+  options = {
+    users.jungleUsers = mkOption {
+      type = types.attrsOf (types.anything // { check = (x: x ? "hosts"); });
+      description = ''
+        Same as users.users but with the extra `hosts` attribute, which controls
+        access to the nodes by `networking.hostName`.
+      '';
+    };
+  };
+
+  config = let
+    allowedUser = host: userConf: builtins.elem host userConf.hosts;
+    filterUsers = host: users: filterAttrs (n: v: allowedUser host v) users;
+    removeHosts = users: mapAttrs (n: v: builtins.removeAttrs v [ "hosts" ]) users;
+    currentHost = config.networking.hostName;
+  in {
+    users.users = removeHosts (filterUsers currentHost config.users.jungleUsers);
+  };
+}

m/module/slurm-client.nix

@@ -0,0 +1,107 @@
+{ config, pkgs, lib, ... }:
+
+let
+  suspendProgram = pkgs.writeScript "suspend.sh" ''
+    #!/usr/bin/env bash
+    exec 1>>/var/log/power_save.log 2>>/var/log/power_save.log
+    set -x
+    export "PATH=/run/current-system/sw/bin:$PATH"
+    echo "$(date) Suspend invoked $0 $*" >> /var/log/power_save.log
+    hosts=$(scontrol show hostnames $1)
+    for host in $hosts; do
+      echo Shutting down host: $host
+      ipmitool -I lanplus -H ''${host}-ipmi -P "" -U "" chassis power off
+    done
+  '';
+
+  resumeProgram = pkgs.writeScript "resume.sh" ''
+    #!/usr/bin/env bash
+    exec 1>>/var/log/power_save.log 2>>/var/log/power_save.log
+    set -x
+    export "PATH=/run/current-system/sw/bin:$PATH"
+    echo "$(date) Suspend invoked $0 $*" >> /var/log/power_save.log
+    hosts=$(scontrol show hostnames $1)
+    for host in $hosts; do
+      echo Starting host: $host
+      ipmitool -I lanplus -H ''${host}-ipmi -P "" -U "" chassis power on
+    done
+  '';
+
+in {
+  systemd.services.slurmd.serviceConfig = {
+    # Kill all processes in the control group on stop/restart. This will kill
+    # all the jobs running, so ensure that we only upgrade when the nodes are
+    # not in use. See:
+    # https://github.com/NixOS/nixpkgs/commit/ae93ed0f0d4e7be0a286d1fca86446318c0c6ffb
+    # https://bugs.schedmd.com/show_bug.cgi?id=2095#c24
+    KillMode = lib.mkForce "control-group";
+  };
+
+  services.slurm = {
+    client.enable = true;
+    controlMachine = "hut";
+    clusterName = "jungle";
+    nodeName = [
+      "owl[1,2]  Sockets=2 CoresPerSocket=14 ThreadsPerCore=2 Feature=owl"
+      "hut       Sockets=2 CoresPerSocket=14 ThreadsPerCore=2"
+    ];
+
+    partitionName = [
+      "owl Nodes=owl[1-2] Default=YES DefaultTime=01:00:00 MaxTime=INFINITE State=UP"
+      "all Nodes=owl[1-2],hut Default=NO DefaultTime=01:00:00 MaxTime=INFINITE State=UP"
+    ];
+
+    # See slurm.conf(5) for more details about these options.
+    extraConfig = ''
+      # Use PMIx for MPI by default. It works okay with MPICH and OpenMPI, but
+      # not with Intel MPI. For that use the compatibility shim libpmi.so
+      # setting I_MPI_PMI_LIBRARY=$pmix/lib/libpmi.so while maintaining the PMIx
+      # library in SLURM (--mpi=pmix). See more details here:
+      # https://pm.bsc.es/gitlab/rarias/jungle/-/issues/16
+      MpiDefault=pmix
+
+      # When a node reboots return that node to the slurm queue as soon as it
+      # becomes operative again.
+      ReturnToService=2
+
+      # Track all processes by using a cgroup
+      ProctrackType=proctrack/cgroup
+
+      # Enable task/affinity to allow the jobs to run in a specified subset of
+      # the resources. Use the task/cgroup plugin to enable process containment.
+      TaskPlugin=task/affinity,task/cgroup
+
+      # Power off unused nodes until they are requested
+      SuspendProgram=${suspendProgram}
+      SuspendTimeout=60
+      ResumeProgram=${resumeProgram}
+      ResumeTimeout=300
+      SuspendExcNodes=hut
+
+      # Turn the nodes off after 1 hour of inactivity
+      SuspendTime=3600
+
+      # Reduce port range so we can allow only this range in the firewall
+      SrunPortRange=60000-61000
+
+      # Use cores as consumable resources. In SLURM terms, a core may have
+      # multiple hardware threads (or CPUs).
+      SelectType=select/cons_tres
+
+      # Ignore memory constraints and only use unused cores to share a node with
+      # other jobs.
+      SelectTypeParameters=CR_Core
+    '';
+  };
+
+  age.secrets.mungeKey = {
+    file = ../../secrets/munge-key.age;
+    owner = "munge";
+    group = "munge";
+  };
+
+  services.munge = {
+    enable = true;
+    password = config.age.secrets.mungeKey.path;
+  };
+}

m/module/slurm-exporter.nix

@@ -0,0 +1,28 @@
+{ config, lib, pkgs, ... }:
+
+# See also: https://github.com/NixOS/nixpkgs/pull/112010
+# And: https://github.com/NixOS/nixpkgs/pull/115839
+
+with lib;
+
+{
+  systemd.services."prometheus-slurm-exporter" = {
+    wantedBy = [ "multi-user.target" ];
+    after = [ "network.target" ];
+    serviceConfig = {
+      Restart = mkDefault "always";
+      PrivateTmp = mkDefault true;
+      WorkingDirectory = mkDefault "/tmp";
+      DynamicUser = mkDefault true;
+      ExecStart = ''
+        ${pkgs.prometheus-slurm-exporter}/bin/prometheus-slurm-exporter --listen-address "127.0.0.1:9341"
+      '';
+      Environment = [
+        "PATH=${pkgs.slurm}/bin"
+        # We need to specify the slurm config to be able to talk to the slurmd
+        # daemon.
+        "SLURM_CONF=${config.services.slurm.etcSlurm}/slurm.conf"
+      ];
+    };
+  };
+}

m/module/slurm-firewall.nix

@@ -0,0 +1,8 @@
+{ ... }:
+
+{
+  networking.firewall = {
+    # Required for PMIx in SLURM, we should find a better way
+    allowedTCPPortRanges = [ { from=1024; to=65535; } ];
+  };
+}

m/module/slurm-hut-nix-store.nix

@@ -0,0 +1,19 @@
+{ ... }:
+
+{
+  # Mount the hut nix store via NFS
+  fileSystems."/mnt/hut-nix-store" = {
+    device = "hut:/nix/store";
+    fsType = "nfs";
+    options = [ "ro" ];
+  };
+
+  systemd.services.slurmd.serviceConfig = {
+    # When running a job, bind the hut store in /nix/store so the paths are
+    # available too.
+    # FIXME: This doesn't keep the programs in /run/current-system/sw/bin
+    # available in the store. Ideally they should be merged but the overlay FS
+    # doesn't work when the underlying directories change.
+    BindReadOnlyPaths = "/mnt/hut-nix-store:/nix/store";
+  };
+}

m/owl1/configuration.nix

@@ -1,7 +1,14 @@
 { config, pkgs, ... }:
 
 {
-  imports = [ ../common/main.nix ];
+  imports = [
+    ../common/xeon.nix
+    ../module/ceph.nix
+    ../module/emulation.nix
+    ../module/slurm-client.nix
+    ../module/slurm-firewall.nix
+    ../module/debuginfod.nix
+  ];
 
   # Select the this using the ID to avoid mismatches
   boot.loader.grub.device = "/dev/disk/by-id/wwn-0x55cd2e414d53566c";

m/owl2/configuration.nix

@@ -1,15 +1,17 @@
-{ config, pkgs, modulesPath, lib, ... }:
+{ config, pkgs, ... }:
 
 {
   imports = [
-    #(modulesPath + "/installer/netboot/netboot-minimal.nix")
-    ../common/main.nix
+    ../common/xeon.nix
+    ../module/ceph.nix
+    ../module/emulation.nix
+    ../module/slurm-client.nix
+    ../module/slurm-firewall.nix
+    ../module/debuginfod.nix
   ];
 
   # Select the this using the ID to avoid mismatches
   boot.loader.grub.device = "/dev/disk/by-id/wwn-0x55cd2e414d535629";
-  #programs.ssh.forwardX11 = false;
-  #programs.ssh.setXAuthLocation = lib.mkForce true;
 
   networking = {
     hostName = "owl2";
@@ -17,6 +19,7 @@
       address = "10.0.40.2";
       prefixLength = 24;
     } ];
+    # Watch out! The OmniPath device is not in the same place here:
     interfaces.ibp129s0.ipv4.addresses = [ {
       address = "10.0.42.2";
       prefixLength = 24;

m/raccoon/configuration.nix

@@ -0,0 +1,64 @@
+{ config, pkgs, lib, modulesPath, ... }:
+
+{
+  imports = [
+    ../common/base.nix
+  ];
+
+  # Don't install Grub on the disk yet
+  boot.loader.grub.device = "nodev";
+
+  # Enable serial console
+  boot.kernelParams = [
+    "console=tty1"
+    "console=ttyS1,115200"
+  ];
+
+  networking = {
+    hostName = "raccoon";
+    # Only BSC DNSs seem to be reachable from the office VLAN
+    nameservers = [ "84.88.52.35" "84.88.52.36" ];
+    defaultGateway = "84.88.51.129";
+    interfaces.eno0.ipv4.addresses = [ {
+      address = "84.88.51.152";
+      prefixLength = 25;
+    } ];
+  };
+
+  # Configure Nvidia driver to use with CUDA
+  hardware.nvidia.package = config.boot.kernelPackages.nvidiaPackages.production;
+  hardware.graphics.enable = true;
+  nixpkgs.config.allowUnfree = true;
+  nixpkgs.config.nvidia.acceptLicense = true;
+  services.xserver.videoDrivers = [ "nvidia" ];
+
+  users.motd = ''
+    ⠀⠀⠀⠀⠀⠀⠀⣀⣀⣄⣠⣀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀
+    ⠀⠀⠀⠀⠀⠀⢰⠇⡀⠀⠙⠻⡿⣦⣀⡀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⡀⠀⠀⠀⠀
+    ⠀⠀⠀⠀⠀⠀⡎⢰⣧⠀⠀⠀⠁⠈⠛⢿⣦⡀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⢀⣠⣴⡦⠶⠟⠓⠚⠻⡄⠀
+    ⠀⠀⠀⠀⠀⠀⣧⠀⣱⣀⣰⣧⠀⢀⠀⣘⣿⣿⣦⣶⣄⣠⡀⠀⠀⣀⣀⣤⣴⣄⣀⣀⡀⠀⠀⠀⠀⠀⠀⠀⠀⢀⣴⣿⠿⠏⠁⠀⣀⣠⣶⣿⡶⣿⠀
+    ⠀⠀⠀⠀⠀⠀⣹⣆⠘⣿⣿⣿⣇⢸⣷⣿⣿⣿⣿⣿⣿⣿⣿⣿⣾⣿⣿⣿⣿⣿⣿⣿⣿⣶⣶⣦⡀⣀⣤⣠⣤⡾⠋⠀⢀⣤⣶⣿⣿⣿⣿⣿⣿⣿⡀
+    ⠀⠀⠀⠀⠀⠀⠘⢿⡄⢼⣿⣿⣿⣿⣿⡟⠻⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣵⣾⡾⠙⣋⣩⣽⣿⣿⣿⣿⢋⡼⠁
+    ⠀⠀⠀⠀⠀⠀⠀⠈⢻⣄⠸⢿⣿⣿⠿⠷⠀⠈⠀⣭⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣷⣾⣿⣿⣿⣿⣿⣿⠇⡼⠁⠀
+    ⠀⠀⠀⠀⠀⠀⠀⠀⢾⣯⡀⠀⢼⡿⠀⠀⠀⢼⠿⣿⣿⣿⣿⣿⣿⣿⣿⣿⡿⣿⡿⣿⣿⣿⠿⣿⣯⣼⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⡿⢋⡼⠁⠀⠀
+    ⠀⠀⠀⠀⠀⠀⠀⠀⠀⢻⡏⠠⣦⠁⠀⠀⠀⠀⠀⠟⠛⠛⣿⣿⣿⣿⣿⠿⠁⠀⠁⢿⠙⠁⠀⠛⠹⣿⣏⣾⣿⣿⣿⣿⣿⣿⣿⣿⠿⠃⣹⠁⠀⠀⠀
+    ⠀⠀⠀⠀⠀⠀⠀⠀⠀⣘⣧⠀⠙⠀⠀⠀⠀⠀⠀⠀⠀⠀⣿⣿⣿⡿⡿⠀⠀⠀⠀⠈⠀⠀⠀⠀⠀⠀⢹⣿⠿⢿⣿⣿⣿⣿⣿⠋⢀⡤⠛⠀⠀⠀⠀
+    ⠀⠀⠀⠀⠀⠀⠀⠀⠀⢹⡯⠁⠀⠀⠀⠀⠀⠀⠀⠀⠀⠸⣿⣿⣿⠇⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠁⠀⢸⣿⣿⣿⠛⠉⠀⣰⠷⠀⠀⠀⠀⠀
+    ⠀⠀⠀⠀⠀⠀⠀⠀⠀⢸⠇⠀⠀⠀⠀⠀⢀⣿⡇⠀⠀⢻⣿⣿⠁⠀⠀⢠⣾⠁⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠸⠟⢿⣿⣄⡀⢸⣿⡀⠀⠀⠀⠀⠀
+    ⠀⠀⠀⠀⠀⠀⠀⠀⢀⣿⠀⠀⠀⢰⣿⣿⡛⣿⣿⡄⢠⡺⠿⡍⠁⢀⣤⣿⣿⣿⠿⣷⣮⣉⣀⡀⠀⠀⠀⠀⠀⠀⠀⠀⠈⣿⠀⠀⠈⣧⠀⠀⠀⠀⠀
+    ⠀⠀⠀⠀⠀⠀⠀⠀⢾⠉⠃⠀⣴⣿⣟⠻⣿⣿⣿⡇⢸⣿⣶⠀⢀⣾⣿⣿⣟⠿⣷⣾⣿⣿⣿⣿⣦⣤⣤⡤⠀⠀⠀⠀⠀⠁⠀⠀⠀⣼⠗⠀⠀⠀⠀
+    ⠀⠀⠐⢄⡀⠀⠀⠀⢘⡀⠀⢶⣾⣿⣿⣿⣿⡿⠋⠁⠈⠻⠉⠀⠚⠻⣿⣿⣿⣶⣾⣿⣿⣿⣿⣿⣿⣷⣬⣤⣶⣦⡀⣾⣶⣇⠀⠀⠈⢉⣷⠀⠀⠀⠀
+    ⠀⠀⠀⠀⠈⠓⠶⢦⡽⠄⣈⣿⣿⣿⣿⣿⠏⠀⠀⠀⠀⠀⠀⠀⠀⠀⠹⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⡓⠙⣿⡟⠀⠀⠀⠈⠛⣷⣶⡄⠀
+    ⠀⠀⠀⠀⠀⠀⠀⢀⣬⠆⢠⣍⣛⠻⣿⡇⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠈⢿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣉⣀⡀⠀⠀⠈⠛⢿⣦⡀
+    ⠐⠒⠒⠶⠶⠶⢦⣬⣟⣥⣀⡉⠛⠻⠶⢁⣤⣾⣿⣿⣿⣷⡄⠀⠀⠀⠀⠀⢸⣿⣿⣿⣿⣿⣟⡛⠿⠭⠭⠭⠭⠭⠿⠿⠿⢿⣿⣟⠃⠀⠀⠀⠹⣟⠓
+    ⠀⣀⣠⠤⠤⢤⣤⣾⣤⡄⣉⣉⣙⣓⡂⣿⣿⣭⣹⣿⣿⣿⣿⡰⣂⣀⢀⠀⠻⣿⠛⠻⠟⠡⣶⣾⣿⣿⣿⣿⣿⣿⣿⡖⠒⠒⠒⠛⠷⢤⡀⢰⣴⣿⡆
+    ⠀⠀⠀⢀⣠⡴⠾⠟⠻⣟⡉⠉⠉⠉⢁⢿⣿⣿⣿⣿⣿⣿⡿⣱⣿⣭⡌⠤⠀⠀⠐⣶⣌⡻⣶⣭⡻⢿⣿⣿⣿⣿⣿⣯⣥⣤⣦⠀⠠⣴⣶⣶⣿⡟⢿
+    ⢀⠔⠊⠉⠀⠀⠀⠀⢸⣯⣤⠀⠀⠠⣼⣮⣟⣿⣿⣿⣻⣭⣾⣿⣿⣷⣶⣦⠶⣚⣾⣿⣿⣷⣜⣿⣿⣶⣝⢿⣿⣿⣿⣿⣷⣦⣄⣰⡄⠈⢿⣿⡿⣇⠀
+    ⠀⠀⠀⠀⠀⠀⠀⠀⠈⢡⢇⠀⠀⣠⣿⣿⣿⣯⣟⣛⣛⣛⣛⣛⣩⣭⣴⣶⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣷⣿⣿⣿⣿⣿⣿⣿⣿⣿⣦⣻⣿⣧⠀⠀
+    ⠀⠀⠀⠀⠀⠀⠀⠀⠀⣾⠏⠀⢹⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣦⣍⣿⣿⣿⣿⡄⠀
+    ⠀⠀⠀⠀⠀⠀⠀⠀⠀⣿⣾⡁⢈⣾⣿⡿⠛⣛⣿⣿⣿⣿ DO YOU BRING FEEDS? ⣿⣿⣿⣿⣿⣿⡏⠈⠙⠈⠁⠀
+    ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠛⡿⠛⠉⣽⣿⣷⣾⡿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⣿⡿⠷⠌⠛⠉⠀⠁⠀⠀⠀⠀⠀
+    ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠈⠀⠀⠹⠋⠀⢻⣿⣿⣿⣿⠿⢿⣿⣿⣿⣿⣿⣿⠿⣿⣿⣿⣿⠿⠛⠋⠉⠁⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀
+    ⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠉⠉⠁⠀⠀⠀⠀⠀⠈⠉⠉⠀⠀⠈⠋⠉⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀ 
+  '';
+}

pkgs/ceph.nix

@@ -1,405 +0,0 @@
-{ lib
-, stdenv
-, runCommand
-, fetchurl
-, fetchFromGitHub
-, fetchPypi
-
-# Build time
-, cmake
-, ensureNewerSourcesHook
-, fmt
-, git
-, makeWrapper
-, nasm
-, pkg-config
-, which
-
-# Tests
-, nixosTests
-
-# Runtime dependencies
-, arrow-cpp
-, babeltrace
-, boost179
-, bzip2
-, cryptsetup
-, cunit
-, doxygen
-, gperf
-, graphviz
-, gtest
-, icu
-, libcap
-, libcap_ng
-, libnl
-, libxml2
-, lttng-ust
-, lua
-, lz4
-, oath-toolkit
-, openldap
-, python310
-, rdkafka
-, rocksdb
-, snappy
-, sqlite
-, utf8proc
-, zlib
-, zstd
-
-# Optional Dependencies
-, curl ? null
-, expat ? null
-, fuse ? null
-, libatomic_ops ? null
-, libedit ? null
-, libs3 ? null
-, yasm ? null
-
-# Mallocs
-, gperftools ? null
-, jemalloc ? null
-
-# Crypto Dependencies
-, cryptopp ? null
-, nspr ? null
-, nss ? null
-
-# Linux Only Dependencies
-, linuxHeaders
-, util-linux
-, libuuid
-, udev
-, keyutils
-, rdma-core
-, rabbitmq-c
-, libaio ? null
-, libxfs ? null
-, liburing ? null
-, zfs ? null
-, ...
-}:
-
-# We must have one crypto library
-assert cryptopp != null || (nss != null && nspr != null);
-
-let
-  shouldUsePkg = pkg: if pkg != null && pkg.meta.available then pkg else null;
-
-  optYasm = shouldUsePkg yasm;
-  optExpat = shouldUsePkg expat;
-  optCurl = shouldUsePkg curl;
-  optFuse = shouldUsePkg fuse;
-  optLibedit = shouldUsePkg libedit;
-  optLibatomic_ops = shouldUsePkg libatomic_ops;
-  optLibs3 = shouldUsePkg libs3;
-
-  optJemalloc = shouldUsePkg jemalloc;
-  optGperftools = shouldUsePkg gperftools;
-
-  optCryptopp = shouldUsePkg cryptopp;
-  optNss = shouldUsePkg nss;
-  optNspr = shouldUsePkg nspr;
-
-  optLibaio = shouldUsePkg libaio;
-  optLibxfs = shouldUsePkg libxfs;
-  optZfs = shouldUsePkg zfs;
-
-  # Downgrade rocksdb, 7.10 breaks ceph
-  rocksdb' = rocksdb.overrideAttrs {
-    version = "7.9.2";
-    src = fetchFromGitHub {
-      owner = "facebook";
-      repo = "rocksdb";
-      rev = "refs/tags/v7.9.2";
-      hash = "sha256-5P7IqJ14EZzDkbjaBvbix04ceGGdlWBuVFH/5dpD5VM=";
-    };
-  };
-
-  hasRadosgw = optExpat != null && optCurl != null && optLibedit != null;
-
-  # Malloc implementation (can be jemalloc, tcmalloc or null)
-  malloc = if optJemalloc != null then optJemalloc else optGperftools;
-
-  # We prefer nss over cryptopp
-  cryptoStr = if optNss != null && optNspr != null then "nss" else
-    if optCryptopp != null then "cryptopp" else "none";
-
-  cryptoLibsMap = {
-    nss = [ optNss optNspr ];
-    cryptopp = [ optCryptopp ];
-    none = [ ];
-  };
-
-  getMeta = description: with lib; {
-     homepage = "https://ceph.io/en/";
-     inherit description;
-     license = with licenses; [ lgpl21 gpl2 bsd3 mit publicDomain ];
-     maintainers = with maintainers; [ adev ak johanot krav ];
-     platforms = [ "x86_64-linux" "aarch64-linux" ];
-   };
-
-  ceph-common = with python.pkgs; buildPythonPackage {
-    pname = "ceph-common";
-    inherit src version;
-
-    sourceRoot = "ceph-${version}/src/python-common";
-
-    propagatedBuildInputs = [
-      pyyaml
-    ];
-
-    nativeCheckInputs = [
-      pytestCheckHook
-    ];
-
-    disabledTests = [
-      # requires network access
-      "test_valid_addr"
-    ];
-
-    meta = getMeta "Ceph common module for code shared by manager modules";
-  };
-
-  # Watch out for python <> boost compatibility
-  python = python310.override {
-    packageOverrides = self: super: {
-      sqlalchemy = super.sqlalchemy.overridePythonAttrs rec {
-        version = "1.4.46";
-        src = fetchPypi {
-          pname = "SQLAlchemy";
-          inherit version;
-          hash = "sha256-aRO4JH2KKS74MVFipRkx4rQM6RaB8bbxj2lwRSAMSjA=";
-        };
-        disabledTestPaths = [
-          "test/aaa_profiling"
-          "test/ext/mypy"
-        ];
-      };
-    };
-  };
-
-  boost = boost179.override {
-    enablePython = true;
-    inherit python;
-  };
-
-  # TODO: split this off in build and runtime environment
-  ceph-python-env = python.withPackages (ps: with ps; [
-    ceph-common
-
-    # build time
-    cython
-
-    # debian/control
-    bcrypt
-    cherrypy
-    influxdb
-    jinja2
-    kubernetes
-    natsort
-    numpy
-    pecan
-    prettytable
-    pyjwt
-    pyopenssl
-    python-dateutil
-    pyyaml
-    requests
-    routes
-    scikit-learn
-    scipy
-    setuptools
-    sphinx
-    virtualenv
-    werkzeug
-
-    # src/pybind/mgr/requirements-required.txt
-    cryptography
-    jsonpatch
-
-    # src/tools/cephfs/shell/setup.py
-    cmd2
-    colorama
-  ]);
-  inherit (ceph-python-env.python) sitePackages;
-
-  version = "18.2.0";
-  src = fetchurl {
-    url = "https://download.ceph.com/tarballs/ceph-${version}.tar.gz";
-    hash = "sha256:0k9nl6xi5brva51rr14m7ig27mmmd7vrpchcmqc40q3c2khn6ns9";
-  };
-in rec {
-  ceph = stdenv.mkDerivation {
-    pname = "ceph";
-    inherit src version;
-
-    nativeBuildInputs = [
-      cmake
-      fmt
-      git
-      makeWrapper
-      nasm
-      pkg-config
-      python
-      python.pkgs.python # for the toPythonPath function
-      python.pkgs.wrapPython
-      which
-      (ensureNewerSourcesHook { year = "1980"; })
-      # for building docs/man-pages presumably
-      doxygen
-      graphviz
-    ];
-
-    enableParallelBuilding = true;
-
-    buildInputs = cryptoLibsMap.${cryptoStr} ++ [
-      arrow-cpp
-      babeltrace
-      boost
-      bzip2
-      ceph-python-env
-      cryptsetup
-      cunit
-      gperf
-      gtest
-      icu
-      libcap
-      libnl
-      libxml2
-      lttng-ust
-      lua
-      lz4
-      malloc
-      oath-toolkit
-      openldap
-      optLibatomic_ops
-      optLibs3
-      optYasm
-      rdkafka
-      rocksdb'
-      snappy
-      sqlite
-      utf8proc
-      zlib
-      zstd
-    ] ++ lib.optionals stdenv.isLinux [
-      keyutils
-      libcap_ng
-      liburing
-      libuuid
-      linuxHeaders
-      optLibaio
-      optLibxfs
-      optZfs
-      rabbitmq-c
-      rdma-core
-      udev
-      util-linux
-    ] ++ lib.optionals hasRadosgw [
-      optCurl
-      optExpat
-      optFuse
-      optLibedit
-    ];
-
-    pythonPath = [ ceph-python-env "${placeholder "out"}/${ceph-python-env.sitePackages}" ];
-
-    preConfigure =''
-      substituteInPlace src/common/module.c --replace "/sbin/modinfo"  "modinfo"
-      substituteInPlace src/common/module.c --replace "/sbin/modprobe" "modprobe"
-      substituteInPlace src/common/module.c --replace "/bin/grep" "grep"
-
-      # install target needs to be in PYTHONPATH for "*.pth support" check to succeed
-      # set PYTHONPATH, so the build system doesn't silently skip installing ceph-volume and others
-      export PYTHONPATH=${ceph-python-env}/${sitePackages}:$lib/${sitePackages}:$out/${sitePackages}
-      patchShebangs src/
-    '';
-
-    cmakeFlags = [
-      "-DCMAKE_INSTALL_DATADIR=${placeholder "lib"}/lib"
-
-      "-DWITH_CEPHFS_SHELL:BOOL=ON"
-      "-DWITH_SYSTEMD:BOOL=OFF"
-      # `WITH_JAEGER` requires `thrift` as a depenedncy (fine), but the build fails with:
-      #     CMake Error at src/opentelemetry-cpp-stamp/opentelemetry-cpp-build-Release.cmake:49 (message):
-      #     Command failed: 2
-      #
-      #        'make' 'opentelemetry_trace' 'opentelemetry_exporter_jaeger_trace'
-      #
-      #     See also
-      #
-      #        /build/ceph-18.2.0/build/src/opentelemetry-cpp/src/opentelemetry-cpp-stamp/opentelemetry-cpp-build-*.log
-      # and that file contains:
-      #     /build/ceph-18.2.0/src/jaegertracing/opentelemetry-cpp/exporters/jaeger/src/TUDPTransport.cc: In member function 'virtual void opentelemetry::v1::exporter::jaeger::TUDPTransport::close()':
-      #     /build/ceph-18.2.0/src/jaegertracing/opentelemetry-cpp/exporters/jaeger/src/TUDPTransport.cc:71:7: error: '::close' has not been declared; did you mean 'pclose'?
-      #       71 |     ::THRIFT_CLOSESOCKET(socket_);
-      #          |       ^~~~~~~~~~~~~~~~~~
-      # Looks like `close()` is somehow not included.
-      # But the relevant code is already removed in `open-telemetry` 1.10: https://github.com/open-telemetry/opentelemetry-cpp/pull/2031
-      # So it's proably not worth trying to fix that for this Ceph version,
-      # and instead just disable Ceph's Jaeger support.
-      "-DWITH_JAEGER:BOOL=OFF"
-      "-DWITH_TESTS:BOOL=OFF"
-
-      # Use our own libraries, where possible
-      "-DWITH_SYSTEM_ARROW:BOOL=ON" # Only used if other options enable Arrow support.
-      "-DWITH_SYSTEM_BOOST:BOOL=ON"
-      "-DWITH_SYSTEM_GTEST:BOOL=ON"
-      "-DWITH_SYSTEM_ROCKSDB:BOOL=ON"
-      "-DWITH_SYSTEM_UTF8PROC:BOOL=ON"
-      "-DWITH_SYSTEM_ZSTD:BOOL=ON"
-
-      # TODO breaks with sandbox, tries to download stuff with npm
-      "-DWITH_MGR_DASHBOARD_FRONTEND:BOOL=OFF"
-      # WITH_XFS has been set default ON from Ceph 16, keeping it optional in nixpkgs for now
-      ''-DWITH_XFS=${if optLibxfs != null then "ON" else "OFF"}''
-    ] ++ lib.optional stdenv.isLinux "-DWITH_SYSTEM_LIBURING=ON";
-
-    postFixup = ''
-      wrapPythonPrograms
-      wrapProgram $out/bin/ceph-mgr --prefix PYTHONPATH ":" "$(toPythonPath ${placeholder "out"}):$(toPythonPath ${ceph-python-env})"
-
-      # Test that ceph-volume exists since the build system has a tendency to
-      # silently drop it with misconfigurations.
-      test -f $out/bin/ceph-volume
-    '';
-
-    outputs = [ "out" "lib" "dev" "doc" "man" ];
-
-    doCheck = false; # uses pip to install things from the internet
-
-    # Takes 7+h to build with 2 cores.
-    requiredSystemFeatures = [ "big-parallel" ];
-
-    meta = getMeta "Distributed storage system";
-
-    passthru = {
-      inherit version;
-      tests = {
-        inherit (nixosTests)
-          ceph-multi-node
-          ceph-single-node
-          ceph-single-node-bluestore;
-      };
-    };
-  };
-
-  ceph-client = runCommand "ceph-client-${version}" {
-      meta = getMeta "Tools needed to mount Ceph's RADOS Block Devices/Cephfs";
-    } ''
-      mkdir -p $out/{bin,etc,${sitePackages},share/bash-completion/completions}
-      cp -r ${ceph}/bin/{ceph,.ceph-wrapped,rados,rbd,rbdmap} $out/bin
-      cp -r ${ceph}/bin/ceph-{authtool,conf,dencoder,rbdnamer,syn} $out/bin
-      cp -r ${ceph}/bin/rbd-replay* $out/bin
-      cp -r ${ceph}/sbin/mount.ceph $out/bin
-      cp -r ${ceph}/sbin/mount.fuse.ceph $out/bin
-      ln -s bin $out/sbin
-      cp -r ${ceph}/${sitePackages}/* $out/${sitePackages}
-      cp -r ${ceph}/etc/bash_completion.d $out/share/bash-completion/completions
-      # wrapPythonPrograms modifies .ceph-wrapped, so lets just update its paths
-      substituteInPlace $out/bin/ceph          --replace ${ceph} $out
-      substituteInPlace $out/bin/.ceph-wrapped --replace ${ceph} $out
-   '';
-}

pkgs/mpich-fix-hwtopo.patch

@@ -0,0 +1,36 @@
+diff --git a/src/util/mpir_hwtopo.c b/src/util/mpir_hwtopo.c
+index 33e88bc..ee3641c 100644
+--- a/src/util/mpir_hwtopo.c
++++ b/src/util/mpir_hwtopo.c
+@@ -200,18 +200,6 @@ int MPII_hwtopo_init(void)
+ #ifdef HAVE_HWLOC
+     bindset = hwloc_bitmap_alloc();
+     hwloc_topology_init(&hwloc_topology);
+-    char *xmlfile = MPIR_pmi_get_jobattr("PMI_hwloc_xmlfile");
+-    if (xmlfile != NULL) {
+-        int rc;
+-        rc = hwloc_topology_set_xml(hwloc_topology, xmlfile);
+-        if (rc == 0) {
+-            /* To have hwloc still actually call OS-specific hooks, the
+-             * HWLOC_TOPOLOGY_FLAG_IS_THISSYSTEM has to be set to assert that the loaded
+-             * file is really the underlying system. */
+-            hwloc_topology_set_flags(hwloc_topology, HWLOC_TOPOLOGY_FLAG_IS_THISSYSTEM);
+-        }
+-        MPL_free(xmlfile);
+-    }
+
+     hwloc_topology_set_io_types_filter(hwloc_topology, HWLOC_TYPE_FILTER_KEEP_ALL);
+     if (!hwloc_topology_load(hwloc_topology)) 
+
+--- a/src/mpi/init/local_proc_attrs.c
++++ b/src/mpi/init/local_proc_attrs.c
+@@ -79,10 +79,6 @@ int MPII_init_local_proc_attrs(int *p_thread_required)
+     /* Set the number of tag bits. The device may override this value. */
+     MPIR_Process.tag_bits = MPIR_TAG_BITS_DEFAULT;
+
+-    char *requested_kinds = MPIR_pmi_get_jobattr("PMI_mpi_memory_alloc_kinds");
+-    MPIR_get_supported_memory_kinds(requested_kinds, &MPIR_Process.memory_alloc_kinds);
+-    MPL_free(requested_kinds);
+-
+     return mpi_errno;
+ }

pkgs/overlay.nix

@@ -1,35 +1,45 @@
 final: prev:
 {
-  bsc = prev.bsc.extend (bscFinal: bscPrev: {
-    # Set MPICH as default
-    mpi = bscFinal.mpich;
+  # Set MPICH as default
+  mpi = final.mpich;
 
-    # Configure the network for MPICH
-    mpich = with final; prev.mpich.overrideAttrs (old: {
-      buildInput = old.buildInputs ++ [
-        libfabric
-        pmix
-      ];
-      configureFlags = [
-        "--enable-shared"
-        "--enable-sharedlib"
-        "--with-pm=no"
-        "--with-device=ch4:ofi"
-        "--with-pmi=pmix"
-        "--with-pmix=${final.pmix}"
-        "--with-libfabric=${final.libfabric}"
-        "--enable-g=log"
-      ] ++ lib.optionals (lib.versionAtLeast gfortran.version "10") [
-        "FFLAGS=-fallow-argument-mismatch" # https://github.com/pmodels/mpich/issues/4300
-        "FCFLAGS=-fallow-argument-mismatch"
-      ];
-    });
+  # Configure the network for MPICH
+  mpich = with final; let
+    # pmix comes with the libraries in .out and headers in .dev
+    pmixAll = symlinkJoin {
+      name = "pmix-all";
+      paths = [ pmix.dev pmix.out ];
+    };
+  in prev.mpich.overrideAttrs (old: {
+    patches = [
+      # See https://github.com/pmodels/mpich/issues/6946
+      ./mpich-fix-hwtopo.patch
+    ];
+    buildInput = old.buildInputs ++ [
+      libfabric
+      pmixAll
+    ];
+    configureFlags = [
+      "--enable-shared"
+      "--enable-sharedlib"
+      "--with-pm=no"
+      "--with-device=ch4:ofi"
+      "--with-pmi=pmix"
+      "--with-pmix=${pmixAll}"
+      "--with-libfabric=${libfabric}"
+      "--enable-g=log"
+    ] ++ lib.optionals (lib.versionAtLeast gfortran.version "10") [
+      "FFLAGS=-fallow-argument-mismatch" # https://github.com/pmodels/mpich/issues/4300
+      "FCFLAGS=-fallow-argument-mismatch"
+    ];
   });
 
-  # Update ceph to 18.2.0 until it lands in nixpkgs, see:
-  # https://github.com/NixOS/nixpkgs/pull/247849
-  inherit (prev.callPackage ./ceph.nix {
-    lua = prev.lua5_4;
-    fmt = prev.fmt_8;
-  }) ceph ceph-client;
+  slurm = prev.slurm.overrideAttrs (old: {
+    patches = (old.patches or []) ++ [
+      # See https://bugs.schedmd.com/show_bug.cgi?id=19324
+      ./slurm-rank-expansion.patch
+    ];
+  });
+
+  prometheus-slurm-exporter = prev.callPackage ./slurm-exporter.nix { };
 }

pkgs/slurm-exporter.nix

@@ -0,0 +1,22 @@
+{ buildGoModule, fetchFromGitHub, lib }:
+
+buildGoModule rec {
+  pname = "prometheus-slurm-exporter";
+  version = "0.20";
+
+  src = fetchFromGitHub {
+    rev = version;
+    owner = "vpenso";
+    repo = pname;
+    sha256 = "sha256-KS9LoDuLQFq3KoKpHd8vg1jw20YCNRJNJrnBnu5vxvs=";
+  };
+
+  vendorHash = "sha256-A1dd9T9SIEHDCiVT2UwV6T02BSLh9ej6LC/2l54hgwI=";
+  doCheck = false;
+
+  meta = with lib; {
+    description = "Prometheus SLURM Exporter";
+    homepage = "https://github.com/vpenso/prometheus-slurm-exporter";
+    platforms = platforms.linux;
+  };
+}

pkgs/slurm-rank-expansion.patch

@@ -0,0 +1,11 @@
+--- a/src/plugins/mpi/pmix/pmixp_dmdx.c	2024-03-15 13:05:24.815313882 +0100
++++ b/src/plugins/mpi/pmix/pmixp_dmdx.c	2024-03-15 13:09:53.936900823 +0100
+@@ -314,7 +314,7 @@ static void _dmdx_req(buf_t *buf, int no
+ 	}
+ 
+ 	nsptr = pmixp_nspaces_local();
+-	if (nsptr->ntasks <= rank) {
++	if ((long) nsptr->ntasks <= (long) rank) {
+ 		char *nodename = pmixp_info_job_host(nodeid);
+ 		PMIXP_ERROR("Bad request from %s: nspace \"%s\" has only %d ranks, asked for %d",
+ 			    nodename, ns, nsptr->ntasks, rank);

secrets/ceph-user.age

@@ -0,0 +1,21 @@
+age-encryption.org/v1
+-> ssh-ed25519 AY8zKw J00a6ZOhkupkhLU5WQ0kD05HEF4KKsSs2hwjHKbnnHU
+J14VoNOCqLpScVO7OLXbqTcLI4tcVUHt5cqY/XQmbGs
+-> ssh-ed25519 sgAamA k8R/bSUdvVmlBI6yHPi5NBQPBGM36lPJwsir8DFGgxE
+4ZKC3gYvic6AVrNGgNjwztbUzhxP8ViX5O3wFo9wlrk
+-> ssh-ed25519 HY2yRg 966xf2fTnA6Wq0uYXbXZQOManqITJcCbQS9LZCGEOh4
+Qg5echQSrzqeDqvaMx+5fqi8XyTjAeCsY/UFJX6YnDs
+-> ssh-ed25519 tcumPQ e0U2okrGIoUpLfPYjIRx1V92rE3hZW13nJef+l3kBQg
+LejAUKBl+tPhwocCF00ZHTzFISnwX8og8GvemiMIcyo
+-> ssh-ed25519 JJ1LWg QkzTsPq9Gdh+FNz/a4bDb9LQOreFyxeTC51UNd1fsj0
+ayrlKenETfQzH1Z9drVEWqszQebicGVJve0/pCnxAE8
+-> ssh-ed25519 CAWG4Q lJLW9+dxvyoD4hYzeXeE/4rzJ6HIeEQOB1+fbhV3xw0
+T2RrVCtTuQvya9HiJB7txk3QGrntpsMX9Tt1cyXoW5E
+-> ssh-ed25519 MSF3dg JOZkFb2CfqWKvZIz7lYxXWgv8iEVDkQF8hInDMZvknc
+MHDWxjUw4dNiC1h4MrU9uKKcI3rwkxABm0+5FYMZkok
+-> ~8m;7f-grease
+lDIullfC98RhpTZ4Mk87Td+VtPmwPdgz+iIilpKugUkmV5r4Uqd7yE+5ArA6ekr/
+G/X4EA
+--- Cz4sv9ZunBcVdZCozdTh1zlg1zIASjk2MjYeYfcN9eA
+ÊN	Å$[H˜ÝQËéŠ
+d£š·'­±ö7…·Í²)ÖØÀÊx9yüÐëE¡þÓM7^Ø[ÐMŽ+É&éâö½$8tM¨Ð²

secrets/gitea-runner-token.age

@@ -0,0 +1,9 @@
+age-encryption.org/v1
+-> ssh-ed25519 HY2yRg DQdgCk16Yu524BsrWVf0krnwWzDM6SeaJCgQipOfwCA
+Ab9ocqra/UWJZI+QGMlxUhBu5AzqfjPgXl+ENIiHYGs
+-> ssh-ed25519 CAWG4Q KF9rGCenb3nf+wyz2hyVs/EUEbsmUs5R+1fBxlCibC8
+7++Kxbr3FHVdVfnFdHYdAuR0Tgfd+sRcO6WRss6LhEw
+-> ssh-ed25519 MSF3dg aUe4DhRsu4X8CFOEAnD/XM/o/0qHYSB522woCaAVh0I
+GRcs5cm2YqA/lGhUtbpboBaz7mfgiLaCr+agaB7vACU
+--- 9Q7Ou+Pxq+3RZilCb2dKC/pCFjZEt4rp5KnTUUU7WJ8
+1¬
Mw4‘Í	ì:Hµ@Á/ägLtMÇ,ßÆ¥ô*¡žzñNV5ˆm‚ÍNŽoÞáj1$÷TøG_³E{Œ%“‰1ǯ‘<>H£îAÛp™

secrets/gitlab-runner-docker-token.age

@@ -0,0 +1,9 @@
+age-encryption.org/v1
+-> ssh-ed25519 HY2yRg WvKK6U1wQtx2pbUDfuaUIXTQiCulDkz7hgUCSwMfMzQ
+jLktUMqKuVxukqzz++pHOKvmucUQqeKYy5IwBma7KxY
+-> ssh-ed25519 CAWG4Q XKGuNNoYFl9bdZzsqYYTY7GsEt5sypLW4R+1uk78NmU
+8dIA2GzRAwTGM5CDHSM2BUBsbXzEAUssWUz2PY2PaTg
+-> ssh-ed25519 MSF3dg T630RsKuZIF/bp+KITnIIWWHsg6M/VQGqbWQZxqT+AA
+SraZcgZJVtmUzHF/XR9J7aK5t5EDNpkC/av/WJUT/G8
+--- /12G8pj9sbs591OM/ryhoLnSWWmzYcoqprk9uN/3g18
+ä·ù
¼Â‡%å]yi"ô<>»LÓâùH`ªa$Æþ)¦9ve<76>.0úmÉK<EFBFBD>vƒÀïu"|1cÞ-%ÔÕ"åWFï¡ÞA«<41>hº$•ºj<eñ¶xÅLx«ç.?œÈâ:L…¬–ƒ,ëu»|³‹F|Õi²äÔ

secrets/jungle-robot-password.age

@@ -0,0 +1,10 @@
+age-encryption.org/v1
+-> ssh-ed25519 HY2yRg 3L1Y5upc5qN6fgiFAox5rD/W8n0eQUv5mT39QAdO5Ac
+XkWsmPmzRgHjsvJgsDKJRgHZ7/sBZFmd1Doppj/y390
+-> ssh-ed25519 CAWG4Q v03Qr+fckdIpsxvQG/viKxlF8WNpO4XUe//QcPzH4k0
+afUwi3ccDCRfUxPDdF7ZkoL+0UX1XwqVtiyabDWjVQk
+-> ssh-ed25519 MSF3dg c2hEUk4LslJpiL7v/4UpT8fK7ZiBJ8+uRhZ/vBoRUDE
+YX9EpnJpHo1eDsZtapTVY6jD+81kb588Oik4NoY9jro
+--- LhUkopNtCsyHCLzEYzBFs+vekOkAR4B3VBaiMF/ZF8w
+oË<EFBFBD>…×à»ÂC‚ßHãáàùýy—LØ”ItMèÕåµI×±sMÆ\Í1-±K”ˆ¤‰G:õ™<02>¦
+ÝgáºÙbpF¼Ó¶Í%Y·

secrets/nix-serve.age

@@ -0,0 +1,12 @@
+age-encryption.org/v1
+-> ssh-ed25519 HY2yRg d144D+VvxhYgKtH//uD2qNuVnYX6bh74YqkyM3ZjBwU
+0IeVmFAf4U8Sm0d01O6ZwJ1V2jl/mSMl4wF0MP5LrIg
+-> ssh-ed25519 CAWG4Q H4nKxue/Cj/3KUF5A+/ygHMjjArwgx3SIWwXcqFtyUo
+4k5NJkLUrueLYiPkr2LAwQLWmuaOIsDmV/86ravpleU
+-> ssh-ed25519 MSF3dg HpgUAFHLPs4w0cdJHqTwf8lySkTeV9O9NnBf49ClDHs
+foPIUUgAYe1YSDy6+aMfjN7xv9xud9fDmhRlIztHoEo
+-> vLkF\<-grease
+3GRT+W8gYSpjl/a6Ix9+g9UJnTpl1ZH/oucfR801vfE8y77DV2Jxz/XJwzxYxKG5
+YEhiTGMNbXw/V7E5aVSz6Bdc
+--- GtiHKCZdHByq9j0BSLd544PhbEwTN138E8TFdxipeiA
+¥¿£‹„ÝG$Sº¼ƒRAæÀ¾Th]nÄ8<C384>,ùHœsÈïÚ=p¼™Ù'»<>ô+ôjõÓõŒ9±)ñ:”)‘¸œYâþÑ8³IØõ8:ol<6F>ë’<1F>åÃZÐæ3–PM”F;ÊrYõ“ÞÛ<1F>$¨­y¸LâÙœ¦ÎœàÕUús16Ǿ¡LŒb÷¨²

secrets/secrets.nix

@@ -0,0 +1,17 @@
+let
+  keys = import ../keys.nix;
+  adminsKeys = builtins.attrValues keys.admins;
+  hut = [ keys.hosts.hut ] ++ adminsKeys;
+  # Only expose ceph keys to safe nodes and admins
+  safe = keys.hostGroup.safe ++ adminsKeys;
+in
+{
+  "gitea-runner-token.age".publicKeys = hut;
+  "gitlab-runner-docker-token.age".publicKeys = hut;
+  "gitlab-runner-shell-token.age".publicKeys = hut;
+  "nix-serve.age".publicKeys = hut;
+  "jungle-robot-password.age".publicKeys = hut;
+
+  "ceph-user.age".publicKeys = safe;
+  "munge-key.age".publicKeys = safe;
+}

web/content/access/index.md

@@ -17,6 +17,6 @@ Then, to request access to the machines we will need some information about you:
 1. The salted hash of your login password, generated with `mkpasswd -m sha-512`
 1. An SSH public key of type Ed25519 (can be generated with `ssh-keygen -t ed25519`)
 
-You can send us both an email at <rodrigo.arias@bsc.es> and
-<aleix.rocanonell@bsc.es> with the details, or directly open a merge request in
-the [jungle repository](https://pm.bsc.es/gitlab/rarias/jungle/).
+Send an email to <jungle@bsc.es> with the details, or directly open a
+merge request in the [jungle
+repository](https://pm.bsc.es/gitlab/rarias/jungle/).

web/content/git/_index.md

@@ -0,0 +1,6 @@
+---
+title: "Git"
+description: "Gitea instance"
+---
+
+If you are reading this page, the proxy to the Gitea service is not working.

web/content/lists/_index.md

@@ -0,0 +1,6 @@
+---
+title: "Lists"
+description: "Mailing lists"
+---
+
+If you are reading this page, the proxy to the public-inbox service is not working.

web/content/owl/_index.md

@@ -8,3 +8,11 @@ description: "Low system noise"
 Much like the silent flight of an owl at night, these nodes are configured to
 minimize the system noise and let programs run undisturbed. The list of nodes is
 `owl[1-2]` and are available for jobs with SLURM.
+
+The contents of the nix store of the hut node is made available in the owl nodes
+when a job is running. This allows jobs to access the same paths that are on hut
+to load dependencies.
+
+For now, only the hut node can be used to build new derivations so that they
+appear in the compute nodes. This applies to the `nix build`, `nix develop` and
+`nix shell` commands.

web/content/posts/2023-09-12/_index.md

@@ -0,0 +1,71 @@
+---
+title: "Update 2023-09-12"
+author: "Rodrigo Arias Mallo"
+date: 2023-09-12
+---
+
+This is a summary of notable changes introduced in the jungle cluster in the
+last months.
+
+### New Ceph filesystem available
+
+We have installed the latest [Ceph filesystem][1] (18.2.0) which stores three
+redundant copies of the data so a failure in one disk doesn't cause data loss.
+It is mounted in /ceph and available for use in the owl1, owl2 and hut
+nodes. For now it provides 2.8 TiB of space and it is expected to
+increase when the last storage node is installed.
+
+[1]: https://en.wikipedia.org/wiki/Ceph_(software)
+
+The throughput is limited by the 1 Gigabit Ethernet speed, but should be
+reasonably fast for most workloads. Here is a test with dd which reaches the
+network limit:
+
+```txt
+hut% dd if=/dev/urandom of=/ceph/rarias/urandom bs=1M count=1024
+1024+0 records in
+1024+0 records out
+1073741824 bytes (1,1 GB, 1,0 GiB) copied, 8,98544 s, 119 MB/s
+```
+
+### SLURM power save
+
+The SLURM daemon has been configured to power down the nodes after one hour of
+idling. When a new job is allocated to a node that is powered off, it is
+automatically turned on and as soon as it becomes available it will execute the
+job. Here is an example with two nodes that boot and execute a simple job that
+shows the date.
+
+```txt
+hut% date; srun -N 2 date
+2023-09-12T17:36:09 CEST
+2023-09-12T17:38:26 CEST
+2023-09-12T17:38:18 CEST
+```
+
+You can expect a similar delay (around 2-3 min) while the nodes are starting.
+Notice that while the nodes are kept on, the delay is not noticeable:
+
+```txt
+hut% date; srun -N 2 date
+2023-09-12T17:40:04 CEST
+2023-09-12T17:40:04 CEST
+2023-09-12T17:40:04 CEST
+```
+
+### Power and temperature monitoring
+
+In the cluster, we monitor the temperature and the power draw of all nodes. This
+allows us to understand which machines are not being used and turn them off to
+save energy that otherwise would be wasted. Here is an example where some nodes
+are powered off to save energy:
+
+![power](./power.png)
+
+We also configured the nodes to work at low CPU frequencies, so the temperature
+is kept low to increase the lifespan of the node components. Towards these
+goals, we have configured two alerts that trigger when the CPUs of a node
+exceeds the limit temperature of 80 °C or when the power draw exceeds 350 W.
+
+By keeping the power consumption and temperatures controlled, we can safely
+incorporate more machines that will only be used on demand.

web/hugo.toml

@@ -2,4 +2,27 @@ baseURL = 'https://jungle.bsc.es/'
 languageCode = 'en-us'
 title = 'The jungle'
 theme = 'PaperMod'
-sectionPagesMenu = "main"
+
+[[menu.main]]
+identifier = "grafana"
+name = "Grafana"
+url = "/grafana/"
+weight = 10
+
+[[menu.main]]
+identifier = "Git"
+name = "Git"
+url = "/git/"
+weight = 20
+
+[[menu.main]]
+identifier = "Lists"
+name = "Lists"
+url = "/lists/"
+weight = 30
+
+[[menu.main]]
+identifier = "Posts"
+name = "Posts"
+url = "/posts/"
+weight = 40

web/themes/PaperMod/layouts/partials/header.html

@@ -100,9 +100,9 @@
                 {{- $separator := or $label_text (not site.Params.disableThemeToggle)}}
                 {{- with site.Home.AllTranslations }}
                 <ul class="lang-switch">
-                    {{- if $separator }}<li>|</li>{{ end }}
                     {{- range . -}}
                     {{- if ne $lang .Lang }}
+                    {{- if $separator }}<li>|</li>{{ end }}
                     <li>
                         <a href="{{- .Permalink -}}" title="{{ .Language.Params.languageAltTitle | default (.Language.LanguageName | emojify) | default (.Lang | title) }}"
                             aria-label="{{ .Language.LanguageName | default (.Lang | title) }}">

web/themes/PaperMod/layouts/partials/post_meta.html

@@ -17,5 +17,5 @@
 {{- end }}
 
 {{- with ($scratch.Get "meta") }}
-{{- delimit . " · " -}}
+{{- delimit . " · " | safeHTML  -}}
 {{- end -}}