59 Commits

Author SHA1 Message Date
4f2422eca2 Create directories in /vault for tent users 2025-06-03 19:10:34 +02:00
ec7c883ae4 Add software RAID in tent using 3 disks 2025-06-03 19:00:26 +02:00
c2c4fa0226 Add access to tent to all hut users too 2025-06-03 18:16:57 +02:00
bcb2bb252d Add hut SSH configuration from outside SSF LAN 2025-06-03 18:16:52 +02:00
e813c811ac Don't use proxy in base preset 2025-06-03 18:05:41 +02:00
da20987f99 Add tent machine from xeon04
We moved the tent machine to the server room in the BSC building and is
now directly connected to the raccoon via NAT.

Fixes: #106
2025-06-03 18:05:34 +02:00
650be88015 Create specific SSF rack configuration
Allow xeon machines to optionally inherit SSF configuration such as the
NFS mount point and the network configuration.
2025-06-03 12:22:07 +02:00
b802a59868 Create tracing group and add arocanon in raccoon
Reviewed-by: Aleix Roca Nonell <aleix.rocanonell@bsc.es>
2025-06-03 11:09:41 +02:00
7247f7e665 Extend perf support in raccoon
Reviewed-by: Aleix Roca Nonell <aleix.rocanonell@bsc.es>
2025-06-03 11:09:30 +02:00
1d555871a5 Enable nixdebuginfod in raccoon
Reviewed-by: Aleix Roca Nonell <aleix.rocanonell@bsc.es>
2025-06-03 10:50:01 +02:00
a2535c996d Make raccoon use performance governor
Reviewed-by: Aleix Roca Nonell <aleix.rocanonell@bsc.es>
2025-06-03 10:45:35 +02:00
37e60afb54 Enable binfmt emulation in raccoon
Reviewed-by: Aleix Roca Nonell <aleix.rocanonell@bsc.es>
2025-06-03 10:45:33 +02:00
3fe138a418 Disable nix garbage collector in raccoon
Reviewed-by: Aleix Roca Nonell <aleix.rocanonell@bsc.es>
2025-06-03 10:45:31 +02:00
4e7a9f7ce4 Add dbautist user to raccoon machine
Reviewed-by: Aleix Roca Nonell <aleix.rocanonell@bsc.es>
2025-06-03 10:45:28 +02:00
a6a1af673a Add node exporter monitoring in raccoon
Reviewed-by: Aleix Roca Nonell <aleix.rocanonell@bsc.es>
2025-06-03 10:45:26 +02:00
2a3a7b2fb2 Allow X11 forwarding via SSH
Reviewed-by: Aleix Roca Nonell <aleix.rocanonell@bsc.es>
2025-06-03 10:45:23 +02:00
b4ab1c836a Enable linger for user rarias
Allows services to run without a login session.

Reviewed-by: Aleix Roca Nonell <aleix.rocanonell@bsc.es>
2025-06-03 10:45:19 +02:00
fb8b4defa7 Only proxy SSH git remotes via hut in xeon
Other machines like raccoon have direct access.

Reviewed-by: Aleix Roca Nonell <aleix.rocanonell@bsc.es>
2025-06-03 10:44:31 +02:00
1bcfbf8cd6 Add machine map file
Documents the location, board and serial numbers so we can track the
machines if they move around. Some information is unkown.

Using the Nix language to encode the machines location and properties
allows us to later use that information in the configuration of the
machines themselves.

Reviewed-by: Aleix Boné <abonerib@bsc.es>
2025-06-02 14:55:58 +02:00
9f43a0e13b Remove fox monitoring via IPMI
We will need to setup an VPN to be able to access fox in its new
location, so for now we simply remove the IPMI monitoring.

Reviewed-by: Aleix Boné <abonerib@bsc.es>
2025-06-02 11:26:53 +02:00
3a3c3050ef Monitor fox, gateway and UPC anella via ICMP
Fox should reply once the machine is connected to the UPC network.
Monitoring also the gateway and UPC anella allows us to estimate if the
whole network is down or just fox.

Reviewed-by: Aleix Boné <abonerib@bsc.es>
2025-06-02 11:26:51 +02:00
4419f68948 Update configuration for UPC network
The fox machine will be placed in the UPC network, so we update the
configuration with the new IP and gateway. We won't be able to reach hut
directly so we also remove the host entry and proxy.

Reviewed-by: Aleix Boné <abonerib@bsc.es>
2025-06-02 11:26:48 +02:00
e51fc9ffa5 Disable home via NFS in fox
It won't be accesible anymore as we won't be in the same LAN.

Reviewed-by: Aleix Boné <abonerib@bsc.es>
2025-06-02 11:26:46 +02:00
2ae9e9b635 Rekey all secrets
Fox is no longer able to use munge or ceph, so we remove the key and
rekey them.

Reviewed-by: Aleix Boné <abonerib@bsc.es>
2025-06-02 11:26:44 +02:00
be77f6a5f5 Rotate fox SSH host key
Prevent decrypting old secrets by reading the git history.

Reviewed-by: Aleix Boné <abonerib@bsc.es>
2025-06-02 11:26:42 +02:00
6316a12a67 Distrust fox SSH key
We no longer will share secrets with fox until we can regain our trust.

Reviewed-by: Aleix Boné <abonerib@bsc.es>
2025-06-02 11:26:38 +02:00
db663913d8 Remove Ceph module from fox
It will no longer be accesible from the UPC.

Reviewed-by: Aleix Boné <abonerib@bsc.es>
2025-06-02 11:26:36 +02:00
b4846b0f6c Remove fox from SLURM
Reviewed-by: Aleix Boné <abonerib@bsc.es>
2025-06-02 11:26:20 +02:00
64a52801ed Remove pam_slurm_adopt from fox
We no longer will be able to use SLURM from jungle.

Reviewed-by: Aleix Boné <abonerib@bsc.es>
2025-06-02 11:26:02 +02:00
7a2f37aaa2 Add UPC temperature sensor monitoring
These sensors are part of their air quality measurements, which just
happen to be very close to our server room.

Reviewed-by: Aleix Boné <abonerib@bsc.es>
2025-05-29 13:01:37 +02:00
aae6585f66 Add meteocat exporter
Allows us to track ambient temperature changes and estimate the
temperature delta between the server room and exterior temperature.
We should be able to predict when we would need to stop the machines due
to excesive temperature as summer approaches.

Reviewed-by: Aleix Boné <abonerib@bsc.es>
2025-05-29 13:01:29 +02:00
1c15e77c83 Add custom nix-daemon exporter
Allows us to see which derivations are being built in realtime. It is a
bit of a hack, but it seems to work. We simply look at the environment
of the child processes of nix-daemon (usually bash) and then look for
the $name variable which should hold the current derivation being
built. Needs root to be able to read the environ file of the different
nix-daemon processes as they are owned by the nixbld* users.

See: https://discourse.nixos.org/t/query-ongoing-builds/23486
Reviewed-by: Aleix Boné <abonerib@bsc.es>
2025-05-29 12:57:07 +02:00
82fc3209de Set keep-outputs to true in all machines
From the documentation of keep-outputs, setting it to true would prevent
the GC from removing build time dependencies:

If true, the garbage collector will keep the outputs of non-garbage
derivations. If false (default), outputs will be deleted unless they are
GC roots themselves (or reachable from other roots).

In general, outputs must be registered as roots separately. However,
even if the output of a derivation is registered as a root, the
collector will still delete store paths that are used only at build time
(e.g., the C compiler, or source tarballs downloaded from the network).
To prevent it from doing so, set this option to true.

See: https://nix.dev/manual/nix/2.24/command-ref/conf-file.html#conf-keep-outputs
Reviewed-by: Aleix Roca Nonell <aleix.rocanonell@bsc.es>
2025-04-22 17:27:37 +02:00
abeab18270 Add raccoon node exporter monitoring
Reviewed-by: Aleix Boné <abonerib@bsc.es>
2025-04-22 14:50:08 +02:00
1985b58619 Increase data retention to 5 years
Now that we have more space, we can extend the retention time to 5 years
to hold the monitoring metrics. For a year we have:

	# du -sh /var/lib/prometheus2
	13G     /var/lib/prometheus2

So we can expect it to increase to about 65 GiB. In the future we may
want to reduce some adquisition frequency.

Reviewed-by: Aleix Boné <abonerib@bsc.es>
2025-04-22 14:50:03 +02:00
44bd061823 Don't forward any docker traffic
Access to the 23080 local port will be done by applying the INPUT rules,
which pass through nixos-fw.

Reviewed-by: Aleix Boné <abonerib@bsc.es>
2025-04-15 14:16:15 +02:00
e8c309f584 Allow traffic from docker to enter port 23080
Before:

  hut% sudo docker run -it --rm alpine /bin/ash -xc 'true | nc -w 3 -v 10.0.40.7 23080'
  + true
  + nc -w 3 -v 10.0.40.7 23080
  nc: 10.0.40.7 (10.0.40.7:23080): Operation timed out

After:

  hut% sudo docker run -it --rm alpine /bin/ash -xc 'true | nc -w 3 -v 10.0.40.7 23080'
  + true
  + nc -w 3 -v 10.0.40.7 23080
  10.0.40.7 (10.0.40.7:23080) open

Fixes: #94
Reviewed-by: Aleix Boné <abonerib@bsc.es>
2025-04-15 14:16:10 +02:00
71ae7fb585 Add bscpm04.bsc.es SSH host and public key
Allows fetching repositories from hut and other machines in jungle
without the need to do any extra configuration.

Reviewed-by: Aleix Boné <abonerib@bsc.es>
2025-04-15 14:15:45 +02:00
8834d561d2 Add nix cache documentation section
Include usage from NixOS and non-NixOS hosts and a test with curl to
ensure it can be reached.

Reviewed-by: Rodrigo Arias Mallo <rodrigo.arias@bsc.es>
2025-04-15 14:08:22 +02:00
29daa3c364 Use hut nix cache in owl1, owl2 and raccoon
For owl1 and owl2 directly connect to hut via LAN with HTTP, but for
raccoon pass via the proxy using jungle.bsc.es with HTTPS. There is no
risk of tampering as packages are signed.

Reviewed-by: Rodrigo Arias Mallo <rodrigo.arias@bsc.es>
2025-04-15 14:08:17 +02:00
9c503fbefb Clean all iptables rules on stop
Prevents the "iptables: Chain already exists." error by making sure that
we don't leave any chain on start. The ideal solution is to use
iptables-restore instead, which will do the right job. But this needs to
be changed in NixOS entirely.

Reviewed-by: Aleix Boné <abonerib@bsc.es>
2025-04-15 14:08:14 +02:00
51b6a8b612 Make nginx listen on all interfaces
Needed for local hosts to contact the nix cache via HTTP directly.
We also allow the incoming traffic on port 80.

Reviewed-by: Aleix Boné <abonerib@bsc.es>
2025-04-15 14:08:07 +02:00
52213d388d Fix nginx /cache regex
`nix-serve` does not handle duplicates in the path:
```
hut$ curl http://127.0.0.1:5000/nix-cache-info
StoreDir: /nix/store
WantMassQuery: 1
Priority: 30
hut$ curl http://127.0.0.1:5000//nix-cache-info
File not found.
```

This meant that the cache was not accessible via:
`curl https://jungle.bsc.es/cache/nix-cache-info` but
`curl https://jungle.bsc.es/cachenix-cache-info` worked.

Reviewed-by: Rodrigo Arias Mallo <rodrigo.arias@bsc.es>
2025-04-15 14:08:04 +02:00
edf744db8d Add new GitLab runner for gitlab.bsc.es
It uses docker based on alpine and the host nix store, so we can perform
builds but isolate them from the system.

Reviewed-by: Aleix Boné <abonerib@bsc.es>
2025-04-08 17:41:18 +02:00
b82894eaec Remove SLURM partition all
We no longer have homogeneous nodes so it doesn't make much sense to
allocate a mix of them.

Reviewed-by: Aleix Boné <abonerib@bsc.es>
2025-04-08 17:15:27 +02:00
1c47199891 Add varcila user to hut and fox
Reviewed-by: Aleix Boné <abonerib@bsc.es>
2025-04-08 17:15:25 +02:00
8738bd4eeb Adjust fox slurm config after disabling SMT
Reviewed-by: Aleix Boné <abonerib@bsc.es>
2025-04-08 17:15:23 +02:00
7699783aac Add abonerib user to fox
Reviewed-by: Aleix Boné <abonerib@bsc.es>
2025-04-08 17:15:21 +02:00
fee1d4da7e Don't move doc in web output
Reviewed-by: Aleix Boné <abonerib@bsc.es>
2025-04-08 17:15:19 +02:00
b77ce7fb56 Add quickstart guide
Reviewed-by: Aleix Boné <abonerib@bsc.es>
2025-04-08 17:15:17 +02:00
b4a12625c5 Reject SSH connections without SLURM allocation
Reviewed-by: Aleix Boné <abonerib@bsc.es>
2025-04-08 17:15:15 +02:00
302106ea9a Add users to fox
Reviewed-by: Aleix Boné <abonerib@bsc.es>
2025-04-08 17:15:13 +02:00
96877de8d9 Add dalvare1 user
Reviewed-by: Aleix Boné <abonerib@bsc.es>
2025-04-08 17:15:11 +02:00
8878985be6 Add fox page in jungle website
Reviewed-by: Aleix Boné <abonerib@bsc.es>
2025-04-08 17:15:08 +02:00
737578db34 Mount NVME disks in /nvme{0,1}
Reviewed-by: Aleix Boné <abonerib@bsc.es>
2025-04-08 17:15:06 +02:00
88555e3f8c Exclude fox from being suspended by slurm
Reviewed-by: Aleix Boné <abonerib@bsc.es>
2025-04-08 17:15:04 +02:00
feb2060be7 Use IPMI host names instead of IP addresses
Reviewed-by: Aleix Boné <abonerib@bsc.es>
2025-04-08 17:15:01 +02:00
00999434c2 Add fox IPMI monitoring
Use agenix to store the credentials safely.

Reviewed-by: Aleix Boné <abonerib@bsc.es>
2025-04-08 17:14:59 +02:00
29d58cc62d Add new fox machine
Reviewed-by: Aleix Boné <abonerib@bsc.es>
2025-04-08 17:14:42 +02:00
57 changed files with 1162 additions and 104 deletions

View File

@@ -18,6 +18,7 @@ in
{
nixosConfigurations = {
hut = mkConf "hut";
tent = mkConf "tent";
owl1 = mkConf "owl1";
owl2 = mkConf "owl2";
eudy = mkConf "eudy";
@@ -25,6 +26,7 @@ in
bay = mkConf "bay";
lake2 = mkConf "lake2";
raccoon = mkConf "raccoon";
fox = mkConf "fox";
};
packages.x86_64-linux = self.nixosConfigurations.hut.pkgs // {

View File

@@ -9,9 +9,11 @@ rec {
koro = "ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIImiTFDbxyUYPumvm8C4mEnHfuvtBY1H8undtd6oDd67 koro";
bay = "ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAICvGBzpRQKuQYHdlUQeAk6jmdbkrhmdLwTBqf3el7IgU bay";
lake2 = "ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAINo66//S1yatpQHE/BuYD/Gfq64TY7ZN5XOGXmNchiO0 lake2";
fox = "ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIDwItIk5uOJcQEVPoy/CVGRzfmE1ojrdDcI06FrU4NFT fox";
};
hostGroup = with hosts; rec {
untrusted = [ fox ];
compute = [ owl1 owl2 ];
playground = [ eudy koro ];
storage = [ bay lake2 ];

View File

@@ -2,7 +2,7 @@
{
imports = [
../common/xeon.nix
../common/ssf.nix
../module/monitoring.nix
];

View File

@@ -23,6 +23,7 @@
trusted-users = [ "@wheel" ];
flake-registry = pkgs.writeText "global-registry.json"
''{"flakes":[],"version":2}'';
keep-outputs = true;
};
gc = {

View File

@@ -8,20 +8,10 @@ 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
# Connect to BSC machines via hut proxy too
Host amdlogin1.bsc.es armlogin1.bsc.es hualogin1.bsc.es glogin1.bsc.es glogin2.bsc.es fpgalogin1.bsc.es
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";
"bscpm04.bsc.es".publicKey = "ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIPx4mC0etyyjYUT2Ztc/bs4ZXSbVMrogs1ZTP924PDgT";
"glogin1.bsc.es".publicKey = "ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIFsHsZGCrzpd4QDVn5xoDOtrNBkb0ylxKGlyBt6l9qCz";
"glogin2.bsc.es".publicKey = "ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIFsHsZGCrzpd4QDVn5xoDOtrNBkb0ylxKGlyBt6l9qCz";
};

View File

@@ -20,6 +20,7 @@
rarias = {
uid = 1880;
isNormalUser = true;
linger = true;
home = "/home/Computational/rarias";
description = "Rodrigo Arias";
group = "Computational";
@@ -39,7 +40,7 @@
home = "/home/Computational/arocanon";
description = "Aleix Roca";
group = "Computational";
extraGroups = [ "wheel" ];
extraGroups = [ "wheel" "tracing" ];
hashedPassword = "$6$hliZiW4tULC/tH7p$pqZarwJkNZ7vS0G5llWQKx08UFG9DxDYgad7jplMD8WkZh5k58i4dfPoWtnEShfjTO6JHiIin05ny5lmSXzGM/";
openssh.authorizedKeys.keys = [
"ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIF3zeB5KSimMBAjvzsp1GCkepVaquVZGPYwRIzyzaCba aleix@bsc"
@@ -55,7 +56,7 @@
home = "/home/Computational/rpenacob";
description = "Raúl Peñacoba";
group = "Computational";
hosts = [ "owl1" "owl2" "hut" ];
hosts = [ "owl1" "owl2" "hut" "tent" ];
hashedPassword = "$6$TZm3bDIFyPrMhj1E$uEDXoYYd1z2Wd5mMPfh3DZAjP7ztVjJ4ezIcn82C0ImqafPA.AnTmcVftHEzLB3tbe2O4SxDyPSDEQgJ4GOtj/";
openssh.authorizedKeys.keys = [
"ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIFYfXg37mauGeurqsLpedgA2XQ9d4Nm0ZGo/hI1f7wwH rpenacob@bsc"
@@ -68,7 +69,7 @@
home = "/home/Computational/anavarro";
description = "Antoni Navarro";
group = "Computational";
hosts = [ "hut" "raccoon" ];
hosts = [ "hut" "tent" "raccoon" "fox" ];
hashedPassword = "$6$QdNDsuLehoZTYZlb$CDhCouYDPrhoiB7/seu7RF.Gqg4zMQz0n5sA4U1KDgHaZOxy2as9pbIGeF8tOHJKRoZajk5GiaZv0rZMn7Oq31";
openssh.authorizedKeys.keys = [
"ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAILWjRSlKgzBPZQhIeEtk6Lvws2XNcYwHcwPv4osSgst5 anavarro@ssfhead"
@@ -81,7 +82,7 @@
home = "/home/Computational/abonerib";
description = "Aleix Boné";
group = "Computational";
hosts = [ "owl1" "owl2" "hut" "raccoon" ];
hosts = [ "owl1" "owl2" "hut" "tent" "raccoon" "fox" ];
hashedPassword = "$6$V1EQWJr474whv7XJ$OfJ0wueM2l.dgiJiiah0Tip9ITcJ7S7qDvtSycsiQ43QBFyP4lU0e0HaXWps85nqB4TypttYR4hNLoz3bz662/";
openssh.authorizedKeys.keys = [
"ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIIIFiqXqt88VuUfyANkZyLJNiuroIITaGlOOTMhVDKjf abonerib@bsc"
@@ -107,16 +108,43 @@
home = "/home/Computational/dbautist";
description = "Dylan Bautista Cases";
group = "Computational";
hosts = [ "hut" ];
hosts = [ "hut" "tent" "raccoon" ];
hashedPassword = "$6$a2lpzMRVkG9nSgIm$12G6.ka0sFX1YimqJkBAjbvhRKZ.Hl090B27pdbnQOW0wzyxVWySWhyDDCILjQELky.HKYl9gqOeVXW49nW7q/";
openssh.authorizedKeys.keys = [
"ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIAb+EQBoS98zrCwnGKkHKwMLdYABMTqv7q9E0+T0QmkS dbautist@bsc-848818791"
];
};
dalvare1 = {
uid = 2758;
isNormalUser = true;
home = "/home/Computational/dalvare1";
description = "David Álvarez";
group = "Computational";
hosts = [ "hut" "tent" "fox" ];
hashedPassword = "$6$mpyIsV3mdq.rK8$FvfZdRH5OcEkUt5PnIUijWyUYZvB1SgeqxpJ2p91TTe.3eQIDTcLEQ5rxeg.e5IEXAZHHQ/aMsR5kPEujEghx0";
openssh.authorizedKeys.keys = [
"ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIGEfy6F4rF80r4Cpo2H5xaWqhuUZzUsVsILSKGJzt5jF dalvare1@ssfhead"
];
};
varcila = {
uid = 5650;
isNormalUser = true;
home = "/home/Computational/varcila";
description = "Vincent Arcila";
group = "Computational";
hosts = [ "hut" "tent" "fox" ];
hashedPassword = "$6$oB0Tcn99DcM4Ch$Vn1A0ulLTn/8B2oFPi9wWl/NOsJzaFAWjqekwcuC9sMC7cgxEVb.Nk5XSzQ2xzYcNe5MLtmzkVYnRS1CqP39Y0";
openssh.authorizedKeys.keys = [
"ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIKGt0ESYxekBiHJQowmKpfdouw0hVm3N7tUMtAaeLejK vincent@varch"
];
};
};
groups = {
Computational = { gid = 564; };
tracing = { };
};
};
}

9
m/common/ssf.nix Normal file
View File

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

View File

@@ -11,7 +11,7 @@
proxy = {
default = "http://hut:23080/";
noProxy = "127.0.0.1,localhost,internal.domain,10.0.40.40";
noProxy = "127.0.0.1,localhost,internal.domain,10.0.40.40,hut";
# Don't set all_proxy as go complains and breaks the gitlab runner, see:
# https://github.com/golang/go/issues/16715
allProxy = null;
@@ -34,37 +34,37 @@
# Node Entry for node: mds01 (ID=72)
10.0.40.40 bay mds01 mds01-eth0
10.0.42.40 bay-ib mds01-ib0
10.0.40.141 bay-ipmi mds01-ipmi0
10.0.40.141 bay-ipmi mds01-ipmi0 mds01-ipmi
# Node Entry for node: oss01 (ID=73)
10.0.40.41 oss01 oss01-eth0
10.0.42.41 oss01-ib0
10.0.40.142 oss01-ipmi0
10.0.40.142 oss01-ipmi0 oss01-ipmi
# Node Entry for node: oss02 (ID=74)
10.0.40.42 lake2 oss02 oss02-eth0
10.0.42.42 lake2-ib oss02-ib0
10.0.40.143 lake2-ipmi oss02-ipmi0
10.0.40.143 lake2-ipmi oss02-ipmi0 oss02-ipmi
# Node Entry for node: xeon01 (ID=15)
10.0.40.1 owl1 xeon01 xeon01-eth0
10.0.42.1 owl1-ib xeon01-ib0
10.0.40.101 owl1-ipmi xeon01-ipmi0
10.0.40.101 owl1-ipmi xeon01-ipmi0 xeon01-ipmi
# Node Entry for node: xeon02 (ID=16)
10.0.40.2 owl2 xeon02 xeon02-eth0
10.0.42.2 owl2-ib xeon02-ib0
10.0.40.102 owl2-ipmi xeon02-ipmi0
10.0.40.102 owl2-ipmi xeon02-ipmi0 xeon02-ipmi
# Node Entry for node: xeon03 (ID=17)
10.0.40.3 xeon03 xeon03-eth0
10.0.42.3 xeon03-ib0
10.0.40.103 xeon03-ipmi0
10.0.40.103 xeon03-ipmi0 xeon03-ipmi
# Node Entry for node: xeon04 (ID=18)
10.0.40.4 xeon04 xeon04-eth0
10.0.42.4 xeon04-ib0
10.0.40.104 xeon04-ipmi0
10.0.40.104 xeon04-ipmi0 xeon04-ipmi
# Node Entry for node: xeon05 (ID=19)
10.0.40.5 koro xeon05 xeon05-eth0
@@ -74,17 +74,17 @@
# Node Entry for node: xeon06 (ID=20)
10.0.40.6 xeon06 xeon06-eth0
10.0.42.6 xeon06-ib0
10.0.40.106 xeon06-ipmi0
10.0.40.106 xeon06-ipmi0 xeon06-ipmi
# Node Entry for node: xeon07 (ID=21)
10.0.40.7 hut xeon07 xeon07-eth0
10.0.42.7 hut-ib xeon07-ib0
10.0.40.107 hut-ipmi xeon07-ipmi0
10.0.40.107 hut-ipmi xeon07-ipmi0 xeon07-ipmi
# Node Entry for node: xeon08 (ID=22)
10.0.40.8 eudy xeon08 xeon08-eth0
10.0.42.8 eudy-ib xeon08-ib0
10.0.40.108 eudy-ipmi xeon08-ipmi0
10.0.40.108 eudy-ipmi xeon08-ipmi0 xeon08-ipmi
'';
};
}

8
m/common/ssf/ssh.nix Normal file
View File

@@ -0,0 +1,8 @@
{
# Connect to intranet git hosts via proxy
programs.ssh.extraConfig = ''
# Connect to BSC machines via hut proxy too
Host amdlogin1.bsc.es armlogin1.bsc.es hualogin1.bsc.es glogin1.bsc.es glogin2.bsc.es fpgalogin1.bsc.es
ProxyCommand nc -X connect -x hut:23080 %h %p
'';
}

View File

@@ -1,9 +1,7 @@
{
# Provides the base system for a xeon node.
# Provides the base system for a xeon node, not necessarily in the SSF rack.
imports = [
./base.nix
./xeon/fs.nix
./xeon/console.nix
./xeon/net.nix
];
}

View File

@@ -2,7 +2,7 @@
{
imports = [
../common/xeon.nix
../common/ssf.nix
#(modulesPath + "/installer/netboot/netboot-minimal.nix")
./kernel/kernel.nix

76
m/fox/configuration.nix Normal file
View File

@@ -0,0 +1,76 @@
{ lib, config, pkgs, ... }:
{
imports = [
../common/base.nix
../common/xeon/console.nix
../module/emulation.nix
];
# Select the this using the ID to avoid mismatches
boot.loader.grub.device = "/dev/disk/by-id/wwn-0x500a07514b0c1103";
# No swap, there is plenty of RAM
swapDevices = lib.mkForce [];
boot.initrd.availableKernelModules = [ "xhci_pci" "ahci" "nvme" "usbhid" "usb_storage" "sd_mod" ];
boot.kernelModules = [ "kvm-amd" ];
hardware.cpu.amd.updateMicrocode = lib.mkDefault config.hardware.enableRedistributableFirmware;
hardware.cpu.intel.updateMicrocode = lib.mkForce false;
networking = {
defaultGateway = "147.83.30.130";
nameservers = [ "8.8.8.8" ];
hostName = "fox";
interfaces.enp1s0f0np0.ipv4.addresses = [
{
# UPC network
# Public IP configuration:
# - Hostname: fox.ac.upc.edu
# - IP: 147.83.30.141
# - Gateway: 147.83.30.130
# - NetMask: 255.255.255.192
# Private IP configuration for BMC:
# - Hostname: fox-ipmi.ac.upc.edu
# - IP: 147.83.35.27
# - Gateway: 147.83.35.2
# - NetMask: 255.255.255.0
address = "147.83.30.141";
prefixLength = 26; # 255.255.255.192
}
];
extraHosts = ''
147.83.30.141 fox.ac.upc.edu
147.83.35.27 fox-ipmi.ac.upc.edu
'';
};
# 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" ];
# Mount NVME disks
fileSystems."/nvme0" = { device = "/dev/disk/by-label/nvme0"; fsType = "ext4"; };
fileSystems."/nvme1" = { device = "/dev/disk/by-label/nvme1"; fsType = "ext4"; };
# Make a /nvme{0,1}/$USER directory for each user.
systemd.services.create-nvme-dirs = let
# Take only normal users in fox
users = lib.filterAttrs (_: v: v.isNormalUser) config.users.users;
commands = lib.concatLists (lib.mapAttrsToList
(_: user: [
"install -d -o ${user.name} -g ${user.group} -m 0755 /nvme{0,1}/${user.name}"
]) users);
script = pkgs.writeShellScript "create-nvme-dirs.sh" (lib.concatLines commands);
in {
enable = true;
wants = [ "local-fs.target" ];
after = [ "local-fs.target" ];
wantedBy = [ "multi-user.target" ];
serviceConfig.ExecStart = script;
};
}

View File

@@ -2,7 +2,7 @@
{
imports = [
../common/xeon.nix
../common/ssf.nix
../module/ceph.nix
../module/debuginfod.nix
@@ -56,6 +56,11 @@
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
'';
# Flush all rules and chains on stop so it won't break on start
extraStopCommands = ''
iptables -F
iptables -X
'';
};
};

View File

@@ -1,8 +1,9 @@
{ pkgs, lib, config, ... }:
{
age.secrets.gitlabRunnerShellToken.file = ../../secrets/gitlab-runner-shell-token.age;
age.secrets.gitlabRunnerDockerToken.file = ../../secrets/gitlab-runner-docker-token.age;
age.secrets.gitlab-pm-shell.file = ../../secrets/gitlab-runner-shell-token.age;
age.secrets.gitlab-pm-docker.file = ../../secrets/gitlab-runner-docker-token.age;
age.secrets.gitlab-bsc-docker.file = ../../secrets/gitlab-bsc-docker-token.age;
services.gitlab-runner = {
enable = true;
@@ -21,21 +22,90 @@
"--docker-network-mode host"
];
environmentVariables = {
https_proxy = "http://localhost:23080";
http_proxy = "http://localhost:23080";
https_proxy = "http://hut:23080";
http_proxy = "http://hut:23080";
};
};
in {
# For pm.bsc.es/gitlab
gitlab-pm-shell = common-shell // {
authenticationTokenConfigFile = config.age.secrets.gitlabRunnerShellToken.path;
authenticationTokenConfigFile = config.age.secrets.gitlab-pm-shell.path;
};
gitlab-pm-docker = common-docker // {
authenticationTokenConfigFile = config.age.secrets.gitlabRunnerDockerToken.path;
authenticationTokenConfigFile = config.age.secrets.gitlab-pm-docker.path;
};
gitlab-bsc-docker = {
# gitlab.bsc.es still uses the old token mechanism
registrationConfigFile = config.age.secrets.gitlab-bsc-docker.path;
tagList = [ "docker" "hut" ];
environmentVariables = {
# We cannot access the hut local interface from docker, so we connect
# to hut directly via the ethernet one.
https_proxy = "http://hut:23080";
http_proxy = "http://hut:23080";
};
executor = "docker";
dockerImage = "alpine";
dockerVolumes = [
"/nix/store:/nix/store:ro"
"/nix/var/nix/db:/nix/var/nix/db:ro"
"/nix/var/nix/daemon-socket:/nix/var/nix/daemon-socket:ro"
];
dockerExtraHosts = [
# Required to pass the proxy via hut
"hut:10.0.40.7"
];
dockerDisableCache = true;
registrationFlags = [
# Increase build log length to 64 MiB
"--output-limit 65536"
];
preBuildScript = pkgs.writeScript "setup-container" ''
mkdir -p -m 0755 /nix/var/log/nix/drvs
mkdir -p -m 0755 /nix/var/nix/gcroots
mkdir -p -m 0755 /nix/var/nix/profiles
mkdir -p -m 0755 /nix/var/nix/temproots
mkdir -p -m 0755 /nix/var/nix/userpool
mkdir -p -m 1777 /nix/var/nix/gcroots/per-user
mkdir -p -m 1777 /nix/var/nix/profiles/per-user
mkdir -p -m 0755 /nix/var/nix/profiles/per-user/root
mkdir -p -m 0700 "$HOME/.nix-defexpr"
mkdir -p -m 0700 "$HOME/.ssh"
cat > "$HOME/.ssh/config" << EOF
Host bscpm04.bsc.es gitlab-internal.bsc.es
User git
ProxyCommand nc -X connect -x hut:23080 %h %p
Host amdlogin1.bsc.es armlogin1.bsc.es hualogin1.bsc.es glogin1.bsc.es glogin2.bsc.es fpgalogin1.bsc.es
ProxyCommand nc -X connect -x hut:23080 %h %p
EOF
cat >> "$HOME/.ssh/known_hosts" << EOF
bscpm04.bsc.es ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIPx4mC0etyyjYUT2Ztc/bs4ZXSbVMrogs1ZTP924PDgT
gitlab-internal.bsc.es ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIF9arsAOSRB06hdy71oTvJHG2Mg8zfebADxpvc37lZo3
EOF
. ${pkgs.nix}/etc/profile.d/nix-daemon.sh
# Required to load SSL certificate paths
. ${pkgs.cacert}/nix-support/setup-hook
'';
environmentVariables = {
ENV = "/etc/profile";
USER = "root";
NIX_REMOTE = "daemon";
PATH = "${config.system.path}/bin:/bin:/sbin:/usr/bin:/usr/sbin";
};
};
};
};
# DOCKER* chains are useless, override at FORWARD and nixos-fw
networking.firewall.extraCommands = ''
# Don't forward any traffic from docker
iptables -I FORWARD 1 -p all -i docker0 -j nixos-fw-log-refuse
# Allow incoming traffic from docker to 23080
iptables -A nixos-fw -p tcp -i docker0 -d hut --dport 23080 -j ACCEPT
'';
#systemd.services.gitlab-runner.serviceConfig.Shell = "${pkgs.bash}/bin/bash";
systemd.services.gitlab-runner.serviceConfig.DynamicUser = lib.mkForce false;
systemd.services.gitlab-runner.serviceConfig.User = "gitlab-runner";

View File

@@ -1,13 +0,0 @@
modules:
default:
collectors:
- bmc
- ipmi
- chassis
lan:
collectors:
- ipmi
- chassis
user: ""
pass: ""

View File

@@ -3,7 +3,10 @@
{
imports = [
../module/slurm-exporter.nix
../module/meteocat-exporter.nix
../module/upc-qaire-exporter.nix
./gpfs-probe.nix
./nix-daemon-exporter.nix
];
age.secrets.grafanaJungleRobotPassword = {
@@ -12,6 +15,8 @@
mode = "400";
};
age.secrets.ipmiYml.file = ../../secrets/ipmi.yml.age;
services.grafana = {
enable = true;
settings = {
@@ -44,7 +49,7 @@
services.prometheus = {
enable = true;
port = 9001;
retentionTime = "1y";
retentionTime = "5y";
listenAddress = "127.0.0.1";
};
@@ -73,8 +78,8 @@
enable = true;
group = "root";
user = "root";
configFile = ./ipmi.yml;
#extraFlags = [ "--log.level=debug" ];
configFile = config.age.secrets.ipmiYml.path;
# extraFlags = [ "--log.level=debug" ];
listenAddress = "127.0.0.1";
};
node = {
@@ -106,6 +111,9 @@
"127.0.0.1:${toString config.services.prometheus.exporters.smartctl.port}"
"127.0.0.1:9341" # Slurm exporter
"127.0.0.1:9966" # GPFS custom exporter
"127.0.0.1:9999" # Nix-daemon custom exporter
"127.0.0.1:9929" # Meteocat custom exporter
"127.0.0.1:9928" # UPC Qaire custom exporter
"127.0.0.1:${toString config.services.prometheus.exporters.blackbox.port}"
];
}];
@@ -161,6 +169,9 @@
"8.8.8.8"
"ssfhead"
"anella-bsc.cesca.cat"
"upc-anella.cesca.cat"
"fox.ac.upc.edu"
"arenys5.ac.upc.edu"
];
}];
relabel_configs = [
@@ -206,7 +217,7 @@
# Sets the "instance" label with the remote host we are querying
source_labels = [ "__param_target" ];
separator = ";";
regex = "(.*)";
regex = "(.*)-ipmi"; # Remove "-ipm̀i" at the end
target_label = "instance";
replacement = "\${1}";
action = "replace";
@@ -248,6 +259,14 @@
module = [ "raccoon" ];
};
}
{
job_name = "raccoon";
static_configs = [
{
targets = [ "127.0.0.1:19002" ]; # Node exporter
}
];
}
];
};
}

View File

@@ -12,16 +12,19 @@ let
installPhase = ''
cp -r public $out
'';
# Don't mess doc/
dontFixup = true;
};
in
{
networking.firewall.allowedTCPPorts = [ 80 ];
services.nginx = {
enable = true;
virtualHosts."jungle.bsc.es" = {
root = "${website}";
listen = [
{
addr = "127.0.0.1";
addr = "0.0.0.0";
port = 80;
}
];

26
m/hut/nix-daemon-builds.sh Executable file
View File

@@ -0,0 +1,26 @@
#!/bin/sh
# Locate nix daemon pid
nd=$(pgrep -o nix-daemon)
# Locate children of nix-daemon
pids1=$(tr ' ' '\n' < "/proc/$nd/task/$nd/children")
# For each children, locate 2nd level children
pids2=$(echo "$pids1" | xargs -I @ /bin/sh -c 'cat /proc/@/task/*/children' | tr ' ' '\n')
cat <<EOF
HTTP/1.1 200 OK
Content-Type: text/plain; version=0.0.4; charset=utf-8; escaping=values
# HELP nix_daemon_build Nix daemon derivation build state.
# TYPE nix_daemon_build gauge
EOF
for pid in $pids2; do
name=$(cat /proc/$pid/environ 2>/dev/null | tr '\0' '\n' | rg "^name=(.+)" - --replace '$1' | tr -dc ' [:alnum:]_\-\.')
user=$(ps -o uname= -p "$pid")
if [ -n "$name" -a -n "$user" ]; then
printf 'nix_daemon_build{user="%s",name="%s"} 1\n' "$user" "$name"
fi
done

View File

@@ -0,0 +1,23 @@
{ pkgs, config, lib, ... }:
let
script = pkgs.runCommand "nix-daemon-exporter.sh" { }
''
cp ${./nix-daemon-builds.sh} $out;
chmod +x $out
''
;
in
{
systemd.services.nix-daemon-exporter = {
description = "Daemon to export nix-daemon metrics";
path = [ pkgs.procps pkgs.ripgrep ];
wantedBy = [ "default.target" ];
serviceConfig = {
Type = "simple";
ExecStart = "${pkgs.socat}/bin/socat TCP4-LISTEN:9999,fork EXEC:${script}";
# Needed root to read the environment, potentially unsafe
User = "root";
Group = "root";
};
};
}

View File

@@ -1,15 +1,15 @@
- targets:
- 10.0.40.101
- 10.0.40.102
- 10.0.40.103
- 10.0.40.104
- 10.0.40.105
- 10.0.40.106
- 10.0.40.107
- 10.0.40.108
- owl1-ipmi
- owl2-ipmi
- xeon03-ipmi
- xeon04-ipmi
- koro-ipmi
- xeon06-ipmi
- hut-ipmi
- eudy-ipmi
# Storage
- 10.0.40.141
- 10.0.40.142
- 10.0.40.143
- bay-ipmi
- oss01-ipmi
- lake2-ipmi
labels:
job: ipmi-lan

View File

@@ -2,7 +2,7 @@
{
imports = [
../common/xeon.nix
../common/ssf.nix
#(modulesPath + "/installer/netboot/netboot-minimal.nix")
../eudy/cpufreq.nix

View File

@@ -2,7 +2,7 @@
{
imports = [
../common/xeon.nix
../common/ssf.nix
../module/monitoring.nix
];

70
m/map.nix Normal file
View File

@@ -0,0 +1,70 @@
{
# In physical order from top to bottom (see note below)
ssf = {
# Switches for Ethernet and OmniPath
switch-C6-S1A-05 = { pos=42; size=1; model="Dell S3048-ON"; };
switch-opa = { pos=41; size=1; };
# SSF login
ssfhead = { pos=39; size=2; label="SSFHEAD"; board="R2208WTTYSR"; contact="operations@bsc.es"; };
# Storage
bay = { pos=38; size=1; label="MDS01"; board="S2600WT2R"; sn="BQWL64850303"; contact="rodrigo.arias@bsc.es"; };
lake1 = { pos=37; size=1; label="OSS01"; board="S2600WT2R"; sn="BQWL64850234"; contact="rodrigo.arias@bsc.es"; };
lake2 = { pos=36; size=1; label="OSS02"; board="S2600WT2R"; sn="BQWL64850266"; contact="rodrigo.arias@bsc.es"; };
# Compute xeon
owl1 = { pos=35; size=1; label="SSF-XEON01"; board="S2600WTTR"; sn="BQWL64954172"; contact="rodrigo.arias@bsc.es"; };
owl2 = { pos=34; size=1; label="SSF-XEON02"; board="S2600WTTR"; sn="BQWL64756560"; contact="rodrigo.arias@bsc.es"; };
xeon03 = { pos=33; size=1; label="SSF-XEON03"; board="S2600WTTR"; sn="BQWL64750826"; contact="rodrigo.arias@bsc.es"; };
# Slot 34 empty
koro = { pos=31; size=1; label="SSF-XEON05"; board="S2600WTTR"; sn="BQWL64954293"; contact="rodrigo.arias@bsc.es"; };
xeon06 = { pos=30; size=1; label="SSF-XEON06"; board="S2600WTTR"; sn="BQWL64750846"; contact="antoni.navarro@bsc.es"; };
hut = { pos=29; size=1; label="SSF-XEON07"; board="S2600WTTR"; sn="BQWL64751184"; contact="rodrigo.arias@bsc.es"; };
eudy = { pos=28; size=1; label="SSF-XEON08"; board="S2600WTTR"; sn="BQWL64756586"; contact="aleix.rocanonell@bsc.es"; };
# 16 KNL nodes, 4 per chassis
knl01_04 = { pos=26; size=2; label="KNL01..KNL04"; board="HNS7200APX"; };
knl05_08 = { pos=24; size=2; label="KNL05..KNL18"; board="HNS7200APX"; };
knl09_12 = { pos=22; size=2; label="KNL09..KNL12"; board="HNS7200APX"; };
knl13_16 = { pos=20; size=2; label="KNL13..KNL16"; board="HNS7200APX"; };
# Slot 19 empty
# EPI (hw team, guessed order)
epi01 = { pos=18; size=1; contact="joan.cabre@bsc.es"; };
epi02 = { pos=17; size=1; contact="joan.cabre@bsc.es"; };
epi03 = { pos=16; size=1; contact="joan.cabre@bsc.es"; };
anon = { pos=14; size=2; }; # Unlabeled machine. Operative
# These are old and decommissioned (off)
power8 = { pos=12; size=2; label="BSCPOWER8N3"; decommissioned=true; };
powern1 = { pos=8; size=4; label="BSCPOWERN1"; decommissioned=true; };
gustafson = { pos=7; size=1; label="gustafson"; decommissioned=true; };
odap01 = { pos=3; size=4; label="ODAP01"; decommissioned=true; };
amhdal = { pos=2; size=1; label="AMHDAL"; decommissioned=true; }; # sic
moore = { pos=1; size=1; label="moore (earth)"; decommissioned=true; };
};
bsc2218 = {
raccoon = { board="W2600CR"; sn="QSIP22500829"; contact="rodrigo.arias@bsc.es"; };
tent = { label="SSF-XEON04"; board="S2600WTTR"; sn="BQWL64751229"; contact="rodrigo.arias@bsc.es"; };
};
upc = {
fox = { board="H13DSG-O-CPU"; sn="UM24CS600392"; prod="AS-4125GS-TNRT"; prod_sn="E508839X5103339"; contact="rodrigo.arias@bsc.es"; };
};
# NOTE: Position is specified in "U" units (44.45 mm) and starts at 1 from the
# bottom. Example:
#
# | ... | - [pos+size] <--- Label in chassis
# +--------+
# | node | - [pos+1]
# | 2U | - [pos]
# +------- +
# | ... | - [pos-1]
#
# NOTE: The board and sn refers to the FRU information (Board Product and
# Board Serial) via `ipmitool fru print 0`.
}

View File

@@ -4,7 +4,7 @@
# Don't add hut as a cache to itself
assert config.networking.hostName != "hut";
{
substituters = [ "https://jungle.bsc.es/cache" ];
substituters = [ "http://hut/cache" ];
trusted-public-keys = [ "jungle.bsc.es:pEc7MlAT0HEwLQYPtpkPLwRsGf80ZI26aj29zMw/HH0=" ];
};
}

View File

@@ -0,0 +1,17 @@
{ config, lib, pkgs, ... }:
with lib;
{
systemd.services."prometheus-meteocat-exporter" = {
wantedBy = [ "multi-user.target" ];
after = [ "network.target" ];
serviceConfig = {
Restart = mkDefault "always";
PrivateTmp = mkDefault true;
WorkingDirectory = mkDefault "/tmp";
DynamicUser = mkDefault true;
ExecStart = "${pkgs.meteocat-exporter}/bin/meteocat-exporter";
};
};
}

View File

@@ -47,8 +47,7 @@ in {
];
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"
"owl Nodes=owl[1-2] Default=YES DefaultTime=01:00:00 MaxTime=INFINITE State=UP"
];
# See slurm.conf(5) for more details about these options.
@@ -91,9 +90,29 @@ in {
# Ignore memory constraints and only use unused cores to share a node with
# other jobs.
SelectTypeParameters=CR_Core
# Required for pam_slurm_adopt, see https://slurm.schedmd.com/pam_slurm_adopt.html
# This sets up the "extern" step into which ssh-launched processes will be
# adopted. Alloc runs the prolog at job allocation (salloc) rather than
# when a task runs (srun) so we can ssh early.
PrologFlags=Alloc,Contain,X11
# LaunchParameters=ulimit_pam_adopt will set RLIMIT_RSS in processes
# adopted by the external step, similar to tasks running in regular steps
# LaunchParameters=ulimit_pam_adopt
SlurmdDebug=debug5
#DebugFlags=Protocol,Cgroup
'';
extraCgroupConfig = ''
CgroupPlugin=cgroup/v2
#ConstrainCores=yes
'';
};
# Place the slurm config in /etc as this will be required by PAM
environment.etc.slurm.source = config.services.slurm.etcSlurm;
age.secrets.mungeKey = {
file = ../../secrets/munge-key.age;
owner = "munge";

View File

@@ -0,0 +1,9 @@
{
programs.ssh.extraConfig = ''
Host ssfhead
HostName ssflogin.bsc.es
Host hut
ProxyJump ssfhead
HostName xeon07
'';
}

View File

@@ -0,0 +1,17 @@
{ config, lib, pkgs, ... }:
with lib;
{
systemd.services."prometheus-upc-qaire-exporter" = {
wantedBy = [ "multi-user.target" ];
after = [ "network.target" ];
serviceConfig = {
Restart = mkDefault "always";
PrivateTmp = mkDefault true;
WorkingDirectory = mkDefault "/tmp";
DynamicUser = mkDefault true;
ExecStart = "${pkgs.upc-qaire-exporter}/bin/upc-qaire-exporter";
};
};
}

View File

@@ -2,7 +2,7 @@
{
imports = [
../common/xeon.nix
../common/ssf.nix
../module/ceph.nix
../module/emulation.nix
../module/slurm-client.nix

View File

@@ -2,7 +2,7 @@
{
imports = [
../common/xeon.nix
../common/ssf.nix
../module/ceph.nix
../module/emulation.nix
../module/slurm-client.nix

View File

@@ -3,7 +3,10 @@
{
imports = [
../common/base.nix
../module/hut-substituter.nix
../module/emulation.nix
../module/debuginfod.nix
../module/ssh-hut-extern.nix
../eudy/kernel/perf.nix
];
# Don't install Grub on the disk yet
@@ -24,8 +27,28 @@
address = "84.88.51.152";
prefixLength = 25;
} ];
interfaces.enp5s0f1.ipv4.addresses = [ {
address = "10.0.44.1";
prefixLength = 24;
} ];
nat = {
enable = true;
internalInterfaces = [ "enp5s0f1" ];
externalInterface = "eno0";
};
hosts = {
"10.0.44.4" = [ "tent" ];
};
};
nix.settings = {
substituters = [ "https://jungle.bsc.es/cache" ];
trusted-public-keys = [ "jungle.bsc.es:pEc7MlAT0HEwLQYPtpkPLwRsGf80ZI26aj29zMw/HH0=" ];
};
# Enable performance governor
powerManagement.cpuFreqGovernor = "performance";
# Configure Nvidia driver to use with CUDA
hardware.nvidia.package = config.boot.kernelPackages.nvidiaPackages.production;
hardware.graphics.enable = true;
@@ -33,6 +56,18 @@
nixpkgs.config.nvidia.acceptLicense = true;
services.xserver.videoDrivers = [ "nvidia" ];
# Disable garbage collection for now
nix.gc.automatic = lib.mkForce false;
services.openssh.settings.X11Forwarding = true;
services.prometheus.exporters.node = {
enable = true;
enabledCollectors = [ "systemd" ];
port = 9002;
listenAddress = "127.0.0.1";
};
users.motd = ''

70
m/tent/configuration.nix Normal file
View File

@@ -0,0 +1,70 @@
{ config, pkgs, lib, ... }:
{
imports = [
../common/xeon.nix
../module/emulation.nix
../module/debuginfod.nix
../module/ssh-hut-extern.nix
];
# Select the this using the ID to avoid mismatches
boot.loader.grub.device = "/dev/disk/by-id/wwn-0x55cd2e414d537675";
networking = {
hostName = "tent";
interfaces.eno1.ipv4.addresses = [
{
address = "10.0.44.4";
prefixLength = 24;
}
];
# Only BSC DNSs seem to be reachable from the office VLAN
nameservers = [ "84.88.52.35" "84.88.52.36" ];
defaultGateway = "10.0.44.1";
};
nix.settings = {
substituters = [ "https://jungle.bsc.es/cache" ];
trusted-public-keys = [ "jungle.bsc.es:pEc7MlAT0HEwLQYPtpkPLwRsGf80ZI26aj29zMw/HH0=" ];
};
services.prometheus.exporters.node = {
enable = true;
enabledCollectors = [ "systemd" ];
port = 9002;
listenAddress = "127.0.0.1";
};
boot.swraid = {
enable = true;
mdadmConf = ''
DEVICE partitions
ARRAY /dev/md0 metadata=1.2 UUID=496db1e2:056a92aa:a544543f:40db379d
MAILADDR root
'';
};
fileSystems."/vault" = {
device = "/dev/disk/by-label/vault";
fsType = "ext4";
};
# Make a /vault/$USER directory for each user.
systemd.services.create-vault-dirs = let
# Take only normal users in tent
users = lib.filterAttrs (_: v: v.isNormalUser) config.users.users;
commands = lib.concatLists (lib.mapAttrsToList
(_: user: [
"install -d -o ${user.name} -g ${user.group} -m 0711 /vault/${user.name}"
]) users);
script = pkgs.writeShellScript "create-vault-dirs.sh" (lib.concatLines commands);
in {
enable = true;
wants = [ "local-fs.target" ];
after = [ "local-fs.target" ];
wantedBy = [ "multi-user.target" ];
serviceConfig.ExecStart = script;
};
}

View File

@@ -0,0 +1,25 @@
{ python3Packages, lib }:
python3Packages.buildPythonApplication rec {
pname = "meteocat-exporter";
version = "1.0";
src = ./.;
doCheck = false;
build-system = with python3Packages; [
setuptools
];
dependencies = with python3Packages; [
beautifulsoup4
lxml
prometheus-client
];
meta = with lib; {
description = "MeteoCat Prometheus Exporter";
platforms = platforms.linux;
};
}

View File

@@ -0,0 +1,54 @@
#!/usr/bin/env python3
import time
from prometheus_client import start_http_server, Gauge
from bs4 import BeautifulSoup
from urllib import request
# Configuration -------------------------------------------
meteo_station = "X8" # Barcelona - Zona Universitària
listening_port = 9929
update_period = 60 * 5 # Each 5 min
# ---------------------------------------------------------
metric_tmin = Gauge('meteocat_temp_min', 'Min temperature')
metric_tmax = Gauge('meteocat_temp_max', 'Max temperature')
metric_tavg = Gauge('meteocat_temp_avg', 'Average temperature')
metric_srad = Gauge('meteocat_solar_radiation', 'Solar radiation')
def update(st):
url = 'https://www.meteo.cat/observacions/xema/dades?codi=' + st
response = request.urlopen(url)
data = response.read()
soup = BeautifulSoup(data, 'lxml')
table = soup.find("table", {"class" : "tblperiode"})
rows = table.find_all('tr')
row = rows[-1] # Take the last row
row_data = []
header = row.find('th')
header_text = header.text.strip()
row_data.append(header_text)
for col in row.find_all('td'):
row_data.append(col.text)
try:
# Sometimes it will return '(s/d)' and fail to parse
metric_tavg.set(float(row_data[1]))
metric_tmax.set(float(row_data[2]))
metric_tmin.set(float(row_data[3]))
metric_srad.set(float(row_data[10]))
#print("ok: temp_avg={}".format(float(row_data[1])))
except:
print("cannot parse row: {}".format(row))
metric_tavg.set(float("nan"))
metric_tmax.set(float("nan"))
metric_tmin.set(float("nan"))
metric_srad.set(float("nan"))
if __name__ == '__main__':
start_http_server(port=listening_port, addr="localhost")
while True:
try:
update(meteo_station)
except:
print("update failed")
time.sleep(update_period)

View File

@@ -0,0 +1,11 @@
#!/usr/bin/env python
from setuptools import setup, find_packages
setup(name='meteocat-exporter',
version='1.0',
# Modules to import from other scripts:
packages=find_packages(),
# Executables
scripts=["meteocat-exporter"],
)

View File

@@ -39,7 +39,21 @@ final: prev:
# See https://bugs.schedmd.com/show_bug.cgi?id=19324
./slurm-rank-expansion.patch
];
# Install also the pam_slurm_adopt library to restrict users from accessing
# nodes with no job allocated.
postBuild = (old.postBuild or "") + ''
pushd contribs/pam_slurm_adopt
make "PAM_DIR=$out/lib/security"
popd
'';
postInstall = (old.postInstall or "") + ''
pushd contribs/pam_slurm_adopt
make "PAM_DIR=$out/lib/security" install
popd
'';
});
prometheus-slurm-exporter = prev.callPackage ./slurm-exporter.nix { };
meteocat-exporter = prev.callPackage ./meteocat-exporter/default.nix { };
upc-qaire-exporter = prev.callPackage ./upc-qaire-exporter/default.nix { };
}

View File

@@ -0,0 +1,24 @@
{ python3Packages, lib }:
python3Packages.buildPythonApplication rec {
pname = "upc-qaire-exporter";
version = "1.0";
src = ./.;
doCheck = false;
build-system = with python3Packages; [
setuptools
];
dependencies = with python3Packages; [
prometheus-client
requests
];
meta = with lib; {
description = "UPC Qaire Prometheus Exporter";
platforms = platforms.linux;
};
}

View File

@@ -0,0 +1,11 @@
#!/usr/bin/env python
from setuptools import setup, find_packages
setup(name='upc-qaire-exporter',
version='1.0',
# Modules to import from other scripts:
packages=find_packages(),
# Executables
scripts=["upc-qaire-exporter"],
)

View File

@@ -0,0 +1,74 @@
#!/usr/bin/env python3
import time
from prometheus_client import start_http_server, Gauge
import requests, json
from datetime import datetime, timedelta
# Configuration -------------------------------------------
listening_port = 9928
update_period = 60 * 5 # Each 5 min
# ---------------------------------------------------------
metric_temp = Gauge('upc_c6_s302_temp', 'UPC C6 S302 temperature sensor')
def genparams():
d = {}
d['topic'] = 'TEMPERATURE'
d['shift_dates_to'] = ''
d['datapoints'] = 301
d['devicesAndColors'] = '1148418@@@#40ACB6'
now = datetime.now()
d['fromDate'] = now.strftime('%d/%m/%Y')
d['toDate'] = now.strftime('%d/%m/%Y')
d['serviceFrequency'] = 'NONE'
# WTF!
for i in range(7):
for j in range(48):
key = 'week.days[{}].hours[{}].value'.format(i, j)
d[key] = 'OPEN'
return d
def measure():
# First we need to load session
s = requests.Session()
r = s.get("https://upc.edu/sirena")
if r.status_code != 200:
print("bad HTTP status code on new session: {}".format(r.status_code))
return
if s.cookies.get("JSESSIONID") is None:
print("cannot get JSESSIONID")
return
# Now we can pull the data
url = "https://upcsirena.app.dexma.com/l_12535/analysis/by_datapoints/data.json"
r = s.post(url, data=genparams())
if r.status_code != 200:
print("bad HTTP status code on data: {}".format(r.status_code))
return
#print(r.text)
j = json.loads(r.content)
# Just take the last one
last = j['data']['chartElementList'][-1]
temp = last['values']['1148418-Temperatura']
return temp
if __name__ == '__main__':
start_http_server(port=listening_port, addr="localhost")
while True:
try:
metric_temp.set(measure())
except:
print("measure failed")
metric_temp.set(float("nan"))
time.sleep(update_period)

Binary file not shown.

Binary file not shown.

View File

@@ -0,0 +1,10 @@
age-encryption.org/v1
-> ssh-ed25519 HY2yRg XPOFoZqY+AnKC77jrgNqAm1ADphurfuhO4NRrfiuUDc
iCfMMpGHyaYHGy6ci8sqjUtcPeteLlyvLGEF79VPOEc
-> ssh-ed25519 CAWG4Q 6OsGrnM+/c5lTN81Rvp166K+ygmSIFeSYzXxYg25KGE
Av1zTw2zK4Gufzti9kQaye7C362GCiDRRHzCqBLR33g
-> ssh-ed25519 MSF3dg 8CHqJ7mEDvjvqbmF+eE6Em1Wi6eHAzEUpiExC1gm7S0
bdwzYHw3RAbdHq+RsiFUP++sQ586VUlSnAzAOhiQUjI
--- gA5XSUfjUBol938sC5DbUf8PvQUIr2pNkS2nL95OF9c
<EFBFBD><EFBFBD><EFBFBD>Ea1G7<47><37>ݩ[R<><52>\{~$Go<47>cQ<63>wKP&<26><><EFBFBD>w<EFBFBD><77><EFBFBD>6]
<EFBFBD><EFBFBD>ѣ<EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>^z<7A>̄ 1k<31><6B><03><><EFBFBD><EFBFBD><EFBFBD>Y<EFBFBD><59>2<15>p<>2<EFBFBD><32>K<EFBFBD><4B><EFBFBD><EFBFBD>nok<>/X<><58>pt''<27><>$0co=<3D><EFBFBD>

View File

@@ -1,9 +1,9 @@
age-encryption.org/v1
-> ssh-ed25519 HY2yRg 0sEIUEJBJQ0k0rBfHaOEbq1pNBqsPin4Xq85v0ds9jY
4wzjLapoOcq53nT2K3hSGED4jTDXci25GLHkl/fL4EI
-> ssh-ed25519 CAWG4Q f68ZbJGwXuCZVnqhwbh+8dh0X/MCdjEd+sVtPyBu/hU
u2TQreyWQvP6PGuwuUNKA/AL68560flqSlaItN3k41I
-> ssh-ed25519 MSF3dg HdrtRW2j7mfkLH0/4aJK5R0cWdjf56HYtEZgzHi9EAs
A6MF6tXmSUq2RF2bpmav0GFTRERwluSZGh2snP/KqkA
--- drsezqi7J/g8gm6N10SkfeAWnYct99WUraB5djLJqpo
g<EFBFBD>
-> ssh-ed25519 HY2yRg pXNTB/ailRwSEJG1pXvrzzpz5HqkDZdWVWnOH7JGeQ4
NzA+2fxfkNRy/u+Zq96A02K1Vxy0ETYZjMkDVTKyCY8
-> ssh-ed25519 CAWG4Q 7CLJWn+EAxoWDduXaOSrHaBFHQ4GIpYP/62FFTj3ZTI
vSYV1pQg2qI2ngCzM0nCZAnqdz1tbT4hM5m+/TyGU2c
-> ssh-ed25519 MSF3dg Akmp4NcZcDuaYHta/Vej6zulNSrAOCd5lmSV+OiBGC4
qTxqVzTyywur+GjtUQdbaIUdH1fqCqPe6qPf8iHRa4w
--- uCKNqD1TmZZThOzlpsecBKx/k+noIWhCVMr/pzNwBr8
r'<27>Ƌs4<15>˺<EFBFBD><41>P<05>L<EFBFBD><4C>7` <09><02><>) H<><48>-<2D>0<EFBFBD>AH<41>5<16><><1F><>L<EFBFBD>Qe<51><65>H2b<32><1D><42>CJG<4A>"-S<><1C>\<5C><> <0C>H<n<>V

View File

@@ -1,9 +1,9 @@
age-encryption.org/v1
-> ssh-ed25519 HY2yRg VY8s9s1zuHOv2axmIacwKg2ozsJnskHTQtslRZ3YI1M
fKkJuydLOzF/ciPYSYu4ziSCozdl6sowvDMYZmxqmHY
-> ssh-ed25519 CAWG4Q 2ARFd/7RWQ/QOk47FnJFChaVBgoV4LE6EA+JHezkXgg
MV4g4Llv8Qcd/wUgJyoNG5AXb6o3aFTrOYGC+lXlSzw
-> ssh-ed25519 MSF3dg SKoxWe8Mi8EkBjkESxStOCI5V4C0KYEXIOx7OdENgTA
p/owKwQ4e4pcGV+hqej2AfPU5QaM2i8VfxhlkjCM4Z4
--- 0VWKU5CQiGbiOtQ2tsZZg88oZm1qcUDEnU5zDTtV+KU
<EFBFBD><EFBFBD>u<>cl<1C><>`<19><06>M<1C><>'Vk6Y<36>!<21>=<3D>L<1E><>y<EFBFBD>-<2D><><EFBFBD>O<EFBFBD>Az<02>ƘVEK<45><4B><EFBFBD>R<EFBFBD>_<08>qL|1V<>[)<29>q<EFBFBD><71><EFBFBD>ƓL<C693><4C>Dy<><79>0<EFBFBD>_<><5F><EFBFBD>q)-T,ƪ<16>_9<5F> <20>?<3F><><EFBFBD><EFBFBD>ib<69>1
-> ssh-ed25519 HY2yRg s6iI9f25xulF4KXt+XY07kXXPKxXo7f2Ql/OTHN55Hk
WO4Fd2H9c+HL3+XhUF3BmEZVILlcchGxSrSmL2OEdGw
-> ssh-ed25519 CAWG4Q TBkdpx8k8K1NvW3wcvaF7omKFwEJ2DxWJp3tIOTjwCA
LcYgWRix23AQnw0OQ7f8+8S3J84CHUElX1vKZSETiLE
-> ssh-ed25519 MSF3dg WzrF8kjTP7BXXDjmUp7kPCKguthAW12RPo6Vy2RMmh4
8C3mT9ktudCTANDxhyNszUkbeDG6X4wOJdx825++dYM
--- /w3YQ2UeTi67H1JR0GsdPz2KoLN2Y7BIZfFY+//AWjY
<EFBFBD>ӣ-`P<>@<40><>ބ<><DE84><EFBFBD>)9<>9l<><6C> <0C><>Zf<5A><66>V?I>΍<>w鉐<77> z40 <20>2{i@<40>Z<EFBFBD><5A>x<EFBFBD><78>AHn<48>%<25><> <0C>ʤ<>/W<><57>Ĕ<0F>l<EFBFBD><6C><EFBFBD><EFBFBD>}<7D>&<08><EFBFBD>(<28><><EFBFBD>K<EFBFBD><4B><EFBFBD>S<EFBFBD><53>o<EFBFBD>z<EFBFBD>=<1F>d

BIN
secrets/ipmi.yml.age Normal file

Binary file not shown.

Binary file not shown.

Binary file not shown.

Binary file not shown.

View File

@@ -9,8 +9,10 @@ in
"gitea-runner-token.age".publicKeys = hut;
"gitlab-runner-docker-token.age".publicKeys = hut;
"gitlab-runner-shell-token.age".publicKeys = hut;
"gitlab-bsc-docker-token.age".publicKeys = hut;
"nix-serve.age".publicKeys = hut;
"jungle-robot-password.age".publicKeys = hut;
"ipmi.yml.age".publicKeys = hut;
"ceph-user.age".publicKeys = safe;
"munge-key.age".publicKeys = safe;

View File

@@ -11,7 +11,7 @@ access to the login machine using a resource petition in the BSC intranet.
Then, to request access to the machines we will need some information about you:
1. Which machines you want access to (hut, owl1, owl2, eudy, koro...)
1. Which machines you want access to ([hut](/hut), [fox](/fox), owl1, owl2, eudy, koro...)
1. Your user name and user id (to match the NFS permissions)
1. Your real name and surname (for identification purposes)
1. The salted hash of your login password, generated with `mkpasswd -m sha-512`

10
web/content/doc/_index.md Normal file
View File

@@ -0,0 +1,10 @@
---
title: "Docs"
description: "Documentation for users of jungle machines"
date: 2023-09-15
---
If this is the first time you use any of the jungle machines with NixOS, follow
the [quick start guide](quickstart).

View File

@@ -0,0 +1,234 @@
---
title: "Quick start"
date: 2023-09-15
---
This documentation will guide you on how to build custom packages of software
and use them in the jungle machines. It has been designed to reduce the friction
from users coming from module systems.
You should be able to access the jungle machines, otherwise [request
access](/access).
## Changes from other HPC machines
Users of other machines have been using the Lmod tool (module load ...) to add
or remove programs from their environment, as well as manually building their
own software for too many years.
While we cannot prevent users from continuing to use this tedious mechanism, we
have designed the jungle machines to be much easier to operate by using the nix
package manager.
### Freedom to install packages
When a user wanted to install a package, it was forced to either do it on its
own directory, or request a system administrator to install it in a shared
directory, so other users can also use that package.
This situation is gone, each user can install any package of software by
themselves, without requiring any other authorization. When two users request
the same package, the same copy will be provided.
A new package will be downloaded if it is available (someone already built it)
or will be built from source on demand.
### No changes over time
All users retain the same versions of the packages they request until they
decide to update them.
## Using nix to manage packages
In this chapter we show how to install packages and enter a development shell to
build new programs from source. The examples are done from the hut machine,
read [this page](/access) to request access.
### Installing binaries
To temporarily install new packages, use:
```text
hut% nix shell jungle#gcc jungle#cowsay jungle#ovni
```
Notice that the packages are described as two parts divided by the `#` symbol.
The first part defines where to take the package from and the second part is
the name of the package. For now we will use `jungle#<package>`. You can find
many more packages here:
<https://search.nixos.org/packages>
You will now enter a new shell, where those requested package **binaries are
available in $PATH**:
```text
hut% cowsay hello world
_____________
< hello world >
-------------
\ ^__^
\ (oo)\_______
(__)\ )\/\
||----w |
|| ||
hut% ovniver
LD_LIBRARY_PATH not set
libovni: build v1.11.0 (a7103f8), dynamic v1.11.0 (a7103f8)
hut% gcc --version
gcc (GCC) 13.3.0
Copyright (C) 2023 Free Software Foundation, Inc.
This is free software; see the source for copying conditions. There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
```
### Building programs
The above method only loads new binaries in the `$PATH`. If we try to build a
program that includes headers or links with a library, it will fail to find
them:
```text
hut$ cat test.c
#include <ovni.h>
int main()
{
ovni_version_check();
return 0;
}
hut% gcc test.c -lovni -o test
test.c:1:10: fatal error: ovni.h: No such file or directory
1 | #include <ovni.h>
| ^~~~~~~~
compilation terminated.
```
We could manually add the full path to the ovni include directory with `-I` and
the libraries with `-L`, but there is a tool that already perform these steps
automatically for us, `nix develop`.
Let's go back to our original shell first, where those packages are not
available anymore:
```
hut% ps
PID TTY TIME CMD
2356260 pts/1 00:00:01 zsh
2457268 pts/1 00:00:00 zsh
2457297 pts/1 00:00:00 ps
hut% exit
hut% ovniver
ovniver: command not found
```
### Creating a flake.nix
To define which packages we want, we will write a small file that list them, a
flake.nix file.
First, we will create a new directory where we are going to be working:
```
hut% mkdir example
hut% cd exmple
```
Then place this flake.nix file:
```nix
{
inputs.jungle.url = "jungle";
outputs = { self, jungle }:
let
pkgs = jungle.outputs.packages.x86_64-linux;
in {
devShells.x86_64-linux.default = pkgs.mkShell {
pname = "devshell";
buildInputs = with pkgs; [
ovni gcc cowsay # more packages here...
];
};
};
}
```
Now enter the shell with:
```
hut% nix develop
warning: creating lock file '/home/Computational/rarias/example/flake.lock':
• Added input 'jungle':
'path:/nix/store/27srv8haj6vv4ywrbmw0a8vds561m8rq-source?lastModified=1739479441&narHash=sha256-Kgjs8SO1w9NbPBu8ghwzCxYJ9kvWpoQOT%2BXwPvA9DcU%3D&rev=76396c0d67ef0cf32377d5c1894bb695293bca9d' (2025-02-13)
• Added input 'jungle/agenix':
'github:ryantm/agenix/f6291c5935fdc4e0bef208cfc0dcab7e3f7a1c41?narHash=sha256-b%2Buqzj%2BWa6xgMS9aNbX4I%2BsXeb5biPDi39VgvSFqFvU%3D' (2024-08-10)
• Added input 'jungle/agenix/darwin':
'github:lnl7/nix-darwin/4b9b83d5a92e8c1fbfd8eb27eda375908c11ec4d?narHash=sha256-gzGLZSiOhf155FW7262kdHo2YDeugp3VuIFb4/GGng0%3D' (2023-11-24)
• Added input 'jungle/agenix/darwin/nixpkgs':
follows 'jungle/agenix/nixpkgs'
• Added input 'jungle/agenix/home-manager':
'github:nix-community/home-manager/3bfaacf46133c037bb356193bd2f1765d9dc82c1?narHash=sha256-7ulcXOk63TIT2lVDSExj7XzFx09LpdSAPtvgtM7yQPE%3D' (2023-12-20)
• Added input 'jungle/agenix/home-manager/nixpkgs':
follows 'jungle/agenix/nixpkgs'
• Added input 'jungle/agenix/nixpkgs':
follows 'jungle/nixpkgs'
• Added input 'jungle/agenix/systems':
'github:nix-systems/default/da67096a3b9bf56a91d16901293e51ba5b49a27e?narHash=sha256-Vy1rq5AaRuLzOxct8nz4T6wlgyUR7zLU309k9mBC768%3D' (2023-04-09)
• Added input 'jungle/bscpkgs':
'git+https://git.sr.ht/~rodarima/bscpkgs?ref=refs/heads/master&rev=6782fc6c5b5a29e84a7f2c2d1064f4bcb1288c0f' (2024-11-29)
• Added input 'jungle/bscpkgs/nixpkgs':
follows 'jungle/nixpkgs'
• Added input 'jungle/nixpkgs':
'github:NixOS/nixpkgs/9c6b49aeac36e2ed73a8c472f1546f6d9cf1addc?narHash=sha256-i/UJ5I7HoqmFMwZEH6vAvBxOrjjOJNU739lnZnhUln8%3D' (2025-01-14)
hut$
```
Notice that long list of messages is Nix creating a new flake.lock file with the
current state of the packages. Next invocations will use the same packages as
described by the lock file.
### Building a program from nix develop
Now let's try again building our test program:
```text
hut$ cat test.c
#include <ovni.h>
int main()
{
ovni_version_check();
return 0;
}
hut$ gcc test.c -o test -lovni
hut$ ldd test
linux-vdso.so.1 (0x00007ffff7fc4000)
libovni.so.1 => /nix/store/sqk972akjv0q8dchn8ccjln2llzyyfd0-ovni-1.11.0/lib/libovni.so.1 (0x00007ffff7fab000)
libc.so.6 => /nix/store/nqb2ns2d1lahnd5ncwmn6k84qfd7vx2k-glibc-2.40-36/lib/libc.so.6 (0x00007ffff7db2000)
/nix/store/nqb2ns2d1lahnd5ncwmn6k84qfd7vx2k-glibc-2.40-36/lib/ld-linux-x86-64.so.2 => /nix/store/nqb2ns2d1lahnd5ncwmn6k84qfd7vx2k-glibc-2.40-36/lib64/ld-linux-x86-64.so.2 (0x00007ffff7fc6000)
hut$ ./test
```
Now the ovni.h header and the libovni library are found and the program is
successfully built, linked and executed.
You can add more packages as needed in your flake.nix:
```nix
buildInputs = with pkgs; [
ovni gcc cowsay # more packages here...
];
```
Make sure you exit the develop shell first, and then enter again with `nix
develop`.
## Remember
- `nix shell` places binaries in the `$PATH`.
- `nix develop` enters a development shell where both binaries and the libraries
and includes are available so you can build new programs.

97
web/content/fox/_index.md Normal file
View File

@@ -0,0 +1,97 @@
---
title: "Fox"
description: "AMD Genoa 9684X with 2 NVIDIA RTX4000 GPUs"
date: 2025-02-12
---
![Fox](fox.jpg)
Picture by [Joanne Redwood](https://web.archive.org/web/20191109175146/https://www.inaturalist.org/photos/6568074),
[CC0](http://creativecommons.org/publicdomain/zero/1.0/deed.en).
The *fox* machine is a big GPU server that is configured to run heavy workloads.
It has two fast AMD CPUs with large cache and 2 reasonable NVIDIA GPUs. Here are
the detailed specifications:
- 2x AMD GENOA X 9684X DP/UP 96C/192T 2.55G 1,150M 400W SP5 3D V-cach
- 24x 32GB DDR5-4800 ECC RDIMM (total 768 GiB of RAM)
- 1x 2.5" SSD SATA3 MICRON 5400 MAX 480GB
- 2x 2.5" KIOXIA CM7-R 1.92TB NVMe GEN5 PCIe 5x4
- 2x NVIDIA RTX4000 ADA Gen 20GB GDDR6 PCIe 4.0
## Access
To access the machine, request a SLURM session from [hut](/hut) using the `fox`
partition:
hut% salloc -p fox
Then connect via ssh:
hut% ssh fox
fox%
Follow [these steps](/access) if you don't have access to hut or fox.
## CUDA
To use CUDA, you can use the following `flake.nix` placed in a new directory to
load all the required dependencies:
```nix
{
inputs.jungle.url = "jungle";
outputs = { jungle, ... }: {
devShell.x86_64-linux = let
pkgs = jungle.nixosConfigurations.fox.pkgs;
in pkgs.mkShell {
name = "cuda-env-shell";
buildInputs = with pkgs; [
git gitRepo gnupg autoconf curl
procps gnumake util-linux m4 gperf unzip
# Cuda packages (more at https://search.nixos.org/packages)
cudatoolkit linuxPackages.nvidia_x11
cudaPackages.cuda_cudart.static
cudaPackages.libcusparse
libGLU libGL
xorg.libXi xorg.libXmu freeglut
xorg.libXext xorg.libX11 xorg.libXv xorg.libXrandr zlib
ncurses5 stdenv.cc binutils
];
shellHook = ''
export CUDA_PATH=${pkgs.cudatoolkit}
export LD_LIBRARY_PATH=/var/run/opengl-driver/lib
export SMS=50
'';
};
};
}
```
Then just run `nix develop` from the same directory:
% mkdir cuda
% cd cuda
% vim flake.nix
[...]
% nix develop
$ nvcc -V
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2024 NVIDIA Corporation
Built on Tue_Feb_27_16:19:38_PST_2024
Cuda compilation tools, release 12.4, V12.4.99
Build cuda_12.4.r12.4/compiler.33961263_0
## Filesystems
The machine has several file systems available.
- `$HOME`: Mounted via NFS across all nodes. It is slow and has low capacity.
Don't abuse.
- `/ceph/home/$USER`: Shared Ceph file system across jungle nodes. Slow but high
capacity. Stores three redundant copies of every file.
- `/nvme{0,1}/$USER`: The two local NVME disks, very fast and large capacity.
- `/tmp`: tmpfs, fast but not backed by a disk. Will be erased on reboot.

BIN
web/content/fox/fox.jpg Normal file

Binary file not shown.

After

Width:  |  Height:  |  Size: 126 KiB

View File

@@ -23,7 +23,7 @@ a good candidate for low noise executions.
We provide a binary cache in `hut`, with the aim of avoiding unnecessary
recompilation of packages.
The cache should contain common packages from bsckpkgs, but we don't provide
The cache should contain common packages from bscpkgs, but we don't provide
any guarantee that of what will be available in the cache, or for how long.
We recommend following the latest version of the `jungle` flake to avoid cache
misses.
@@ -38,7 +38,7 @@ enable it for all builds in the system.
```nix
{ ... }: {
nix.settings = {
substituters = [ "https://jungle.bsc.es" ];
substituters = [ "https://jungle.bsc.es/cache" ];
trusted-public-keys = [ "jungle.bsc.es:pEc7MlAT0HEwLQYPtpkPLwRsGf80ZI26aj29zMw/HH0=" ];
};
}
@@ -53,13 +53,15 @@ The cache can also be specified in a per-command basis through the flags
nix build --substituters "https://jungle.bsc.es/cache" --trusted-public-keys "jungle.bsc.es:pEc7MlAT0HEwLQYPtpkPLwRsGf80ZI26aj29zMw/HH0=" <...>
```
Note: you'll have to be a trusted user.
### Nix configuration file (non-nixos)
If using nix outside of NixOS, you'll have to update `nix.conf`
If using nix outside of NixOS, you'll have to update `/etc/nix/nix.conf`
```bash
echo "substituters = https://jungle.bsc.es" >> /etc/nix/nix.conf
echo "trusted-public-keys = jungle.bsc.es:pEc7MlAT0HEwLQYPtpkPLwRsGf80ZI26aj29zMw/HH0=" >> /etc/nix/nix.conf
```
# echo "substituters = https://jungle.bsc.es/cache" >> /etc/nix/nix.conf
# echo "trusted-public-keys = jungle.bsc.es:pEc7MlAT0HEwLQYPtpkPLwRsGf80ZI26aj29zMw/HH0=" >> /etc/nix/nix.conf
```
### Hint in flakes

View File

@@ -3,32 +3,38 @@ languageCode = 'en-us'
title = 'The jungle'
theme = 'PaperMod'
[[menu.main]]
identifier = "doc"
name = "Docs"
url = "/doc/"
weight = 10
[[menu.main]]
identifier = "grafana"
name = "Grafana"
url = "/grafana/"
weight = 10
weight = 20
[[menu.main]]
identifier = "Git"
name = "Git"
url = "/git/"
weight = 20
weight = 30
[[menu.main]]
identifier = "Lists"
name = "Lists"
url = "/lists/"
weight = 30
weight = 40
[[menu.main]]
identifier = "Paste"
name = "Paste"
url = "/paste/"
weight = 40
weight = 50
[[menu.main]]
identifier = "Posts"
name = "Posts"
url = "/posts/"
weight = 50
weight = 60