Application Services > Linux Compute Servers

Linux Compute Servers

What are compute servers?

We have a number of general usage and bookable compute servers that are administered by CSLab. Their purpose is to support compute-intensive research applications, such as Matlab.

Please do not run applications such as web browsers or email clients on the compute servers - general-purpose computing should be done on your desktop if you have one, or on an application server if you use an X terminal of some sort or are connecting remotely. Likewise, compute-intensive processes should not be run on the application servers as they are meant for general-purpose computing.

Although most software on the compute servers should be available as part of the packages provided by the OS vendor, some software which is provided by alternate sources resides in /opt. Matlab is one such piece of software, and there will be multiple version available in /opt as they are released. You should consider setting your PATH or using aliases to make running software in /opt easier; your Point of Contact (PoC) can assist you with this.

If a desired (but not installed) package is available as part of Ubuntu 14.04 LTS, contact your Point of Contact (PoC) and have them put in a request that it be added to core software. Unfortunately, due to the need to keep the software on core servers manageable, such requests may need to be reviewed by the departmental computing committee (DCC).

In order to keep the computational resources of compute servers available for the computationally intensive tasks that they are designed to support, only software appropriate for compute servers will be installed on those servers. For productivity applications such as email readers and web browsers, please use the linux application servers instead.

With respect to software that is neither available via the OS distribution-provided package list nor in /opt, you or your Point of Contact (PoC) can install programs in your home directory or a suitable work partition for use on these servers, so long as the programs do not require Administrator/root privileges to install or run.

General Usage Compute Servers

comps0 and comps1 are Dell Poweredge R210II servers with a quad-core Intel E3-1230v2 CPU. comps2 is a Silicon Mechanics Rackform iServ R331.v4 with two 12-core Intel E5-2697v2 CPUs. comps3 is a Dell PowerEdge 2950 server with dual quad-core Intel X5355 CPUs. comps4 is a SunFire X2270 with two six-core Intel X5675 CPUs. comps0, comps1 and comps3 have 32G, comps2 has 128G and comps4 has 24G. The comps servers are running Ubuntu 14.04 LTS in 64-bit.

Bookable Compute Servers

There are also several compute servers that may be booked by any Point of Contact (PoC) on behalf of a research faculty member or group. The five short term bookable compute servers (compsbk0 and compsbk[2-5]) can be booked for a week at a time. The two long term bookable compute servers (compsltbk0 and compsltbk1) can be booked for a month at a time.

The page Status of CSLab's bookable compute servers has information on the current and future reservations for each server. Please contact your Point of Contact (PoC) for more information or if you wish to book any of the bookable servers.

Short Term Bookable Compute Servers

compsbk0 is a Dell Poweredge 1950 with two quad-core Intel CPUs and 28GB of RAM, running 64-bit Ubuntu 14.04 (i.e. it is identical to comps1 except that it has 28GB of RAM instead of 16GB).

compsbk2 is a Proliant DL980 G7 with eight eight-core Intel X6550 processors (64 cores total) and 256G of RAM, running 64-bit Ubuntu 14.04.

compsbk3 is intended primarily for GPU computing. It has two eight-core AMD Opteron 6128 processors (16 cores total) and 64G of RAM. It is running 64-bit Ubuntu 14.04. compsbk3 has two GPUs: NVIDIA Tesla C2075 with 6GB and EVGA GeForce GTX580 with 3GB. Information on CUDA is provided here.

compsbk4 and compsbk5 are also intended primarily for GPU computing. Both have two six-core Intel E5-2620 processors (12 cores total) and 256G of RAM. compsbk 4 and compsbk5 are running 64-bit Ubuntu 14.04. They have four GPUs: NVIDIA Tesla Fermi M2090 cGPUs with 6GB. Information on CUDA is provided here.

Long Term Bookable Compute Servers

compsltbk0 and compsltbk1 are SunFire X4150 servers with two quad-core Intel Xeon E5440 CPUs and 32GB of RAM, running 64-bit Ubuntu 14.04.

Specifications

In a nutshell, the benchmarks given are calculated by spec.org and compare two cases:

• Integer calculations running multiple jobs simultaneously (SPEC CPU Integer 2006 Rate)

  Floating point calculations running multiple jobs simultaneously (SPEC CPU FP 2006 Rate)

You can find more details on the specifics of these benchmarks at spec.org cpu 2006 performance evaluations and spec.org cpu benchmark details.

RSA key fingerprints

Below are the RSA key fingerprints for the new compute servers so that, when you connect to them via SSH, you can cross-reference the RSA key fingerprint that is reported by the remote host with the ones detailed below, and thus be certain that you are indeed connecting to your intended machine.

 
 comps0.cs.toronto.edu: 0d:2c:31:5c:c1:8c:c7:7b:c9:45:0c:95:38:5e:6b:ce
 comps1.cs.toronto.edu: 02:eb:56:58:2b:1f:e9:29:75:10:2a:e3:1d:7e:5f:a3
 comps2.cs.toronto.edu: da:b3:94:54:16:d8:09:e5:31:7c:48:e5:85:9e:11:4a
 comps3.cs.toronto.edu: de:60:c9:f7:66:57:35:c8:5a:98:33:67:f7:51:35:97
 comps4.cs.toronto.edu: 7f:3d:1b:6b:8e:9f:79:56:cc:c2:30:d1:b2:57:b7:27
 compsbk0.cs.toronto.edu: cc:85:76:a5:4d:70:cd:0b:b8:08:57:e3:6b:10:94:96
 compsbk2.cs.toronto.edu: ca:0f:ce:ba:84:92:f3:a7:be:bc:c6:34:04:1c:4a:55
 compsbk3.cs.toronto.edu: f3:72:b9:6d:1d:1b:09:4b:dc:b6:70:ff:a2:c6:a1:20
 compsbk4.cs.toronto.edu: b7:8e:4f:98:67:18:41:77:5b:49:62:01:53:20:46:e6
 compsbk5.cs.toronto.edu: 74:64:3b:f8:91:41:ff:0e:7a:6f:0a:08:12:51:11:b2
 compsltbk0.cs.toronto.edu: a4:ec:6f:af:5c:64:b8:b1:8d:c8:2f:a2:d9:2c:ce:ac
 compsltbk1.cs.toronto.edu: de:7b:26:c9:4e:2a:58:62:3e:23:2e:64:87:dd:37:7d