A new SPARC64 machine running Debian is available. It has a relatively old T3 processor, but provides 32 CPU cores and 256 threads.
It is already possible to login to this new machine with:
Many thanks to ylibc.org for making this machine available!
The Compile Farm AIX 7.2 system (gcc119) will be offline and unavailable on Monday, May 18, to transfer the system to a different partition. All user files should be available on the new partition when the transfer is complete and the system is restarted. The system will be transferred back to the original partition at a later time.
We have a new machine in the compile farm: gcc203. It is a POWER8 system in big-endian mode running Linux, that is, powerpc64-linux.
It runs Debian; it is a 8-cpu partition of a bigger machine (an S822).
That means you shouldn't normally use more than -j32 (and always much less for non-interactive runs, of course).
The Power ISA machines we now have are:
|Name||Arch||OS||CPU cores||CPU threads||RAM||Disk|
|gcc110||POWER7 BE||Linux||16 cores||64 threads||64 GB||1.8 TB|
|gcc203||POWER8 BE||Linux||8 cores||64 threads||64 GB||1.0 TB|
|gcc112||POWER8 LE||Linux||20 cores||160 threads||256 GB||1.8 TB|
|gcc135||POWER9 LE||Linux||32 cores||128 threads||256 GB||24.9 TB|
|gcc111||POWER7 BE||AIX||12 cores||48 threads||128 GB|
|gcc119||POWER8 BE||AIX||16 cores||128 threads||160 GB|
So, clearly, as a workhorse machine you should use gcc135!
The list of machines currently has some of the CPU/core numbers wrong for gcc203; we're working on it, various programs do not know how to handle virtualised systems well.
Enjoy, and please be mindful of others!
Following a 4-months outage period, gcc20 has been brought back online thanks to our host Inria.
We took this opportunity to perform a much-needed upgrade from wheezy to stretch.
User data has been kept through the upgrade, but please use this opportunity to clean up your old data! Currently, 729 GB of disk space is used out of a 825 GB capacity.
Two farm servers, gcc75 and gcc76, are going to be shut down and moved on May 6th. If you have data you care about on these machines, please fetch it now! The servers may be hosted somewhere else in the future, but there is no ETA.
Many thanks to INSA Rouen for the 7 years of hosting!
We are pleased to announce that a new machine is available in the farm:
gcc220, a physical server running OpenBSD. With 8 cores and 8 GB of RAM, it is more powerful than the existing OpenBSD VM,
Many thanks to openbsd.amsterdam for providing both the machine and the hosting!
The new machine can be reached over SSH at
gcc220.fsffrance.org and is located in Amsterdam.
In November 2018, gcc118 suffered from a filesystem issue after upgrading the kernel.
Thanks to the excellent support from OSUOSL students, the machine has been reinstalled with a more recent OpenSUSE version (15.0) and is available again for farm usage! Unfortunately, we couldn't recover any user data.
The SSH key of the server has changed. You will need to clear the old key from your
ssh-keygen -R gcc118.fsffrance.org
ssh-keygen -R 220.127.116.11
We will look at a way to backup the server SSH keys on all machines to avoid this kind of inconvenience in the future.
We have been monitoring most farm machines with munin for some time. This allows anyone to check whether a machine is heavily loaded or not before starting to use it, but it is time-consuming to check many graphs from several machines.
To quickly see the load of a machine, we now display a usage bar directly in the list of machines! We currently show 3 metrics: CPU usage, memory usage, and disk I/O load. The values are based on the last 48 hours of munin data and should give a good overview of current usage and expected performance.
Nethertheless, these values are indicative and only reflect the average usage. If you plan to use a machine for heavy tasks, you should check the munin graphs to better understand the usage pattern and make sure you don't disrupt the tasks of existing users. In addition, a few values are missing or incorrect: this happens either because the machine was not recently monitored by munin, or because the total number of CPU is incorrectly detected.
The value displayed in the usage bars is computed as a weighted average that gives more importance to high usage values. For instance:
In each case, the arithmetic average would be the same (20% usage). However, we consider that the machine is more loaded in the third case. Intuitively, if you would like to run a new task, using a machine that is already 100% busy (even only 20% of the time) is generally a bad idea: the new task might significantly interfere with the existing ones. In contrast, a machine that is 20% busy still has a lot of room for more tasks.
We are happy to announce that a new IBM server has been added to the farm, thanks to OSUOSL!
This POWER9 server extends the already available IBM hardware in the compile farm: POWER7 (gcc110, gcc111) and POWER8 (gcc112, gcc119). It features 128 CPU cores and 256 GB of RAM, as well as a high-performance I/O setup with 8 disks in a RAID5 configuration. Disk I/O was the main bottleneck for the existing POWER machines, so hopefully this new machine will provide much better performance in practice.
It is already possible to connect to this new machine through SSH at gcc135.fsffrance.org.
gcc123 recently suffered from a disk failure, just a few months after being operational. Today, OSUOSL changed the disk and reinstalled the machine. Because there is no RAID on this machine, any user data stored on the machine was lost.
This is a good opportunity to remind that the compile farm is "best effort": we make no promise whatsover about data integrity, so make sure to always have a copy of your important work somewhere else!
The SSH host keys of gcc123 have been restored, so SSH access should work just fine with the reinstalled system.