Difference between revisions of "Systems"

From Research Computing Center Wiki
Jump to navigation Jump to search
 
(28 intermediate revisions by 3 users not shown)
Line 30: Line 30:
 
===  Sapelo2 ===
 
===  Sapelo2 ===
  
−
Sapelo2 is a Linux cluster that runs a 64-bit CentOS 7.5 operating system and it is managed using Foreman and Puppet. Two physical login nodes are available, with Intel Xeon E5-2680 v3 (Haswell) processors and 128GB of RAM and 24 cores per node.  
+
Sapelo2 is a Linux cluster that runs a 64-bit Rocky 8.8 operating system and it is managed using Warewulf. Several virtual login nodes are available, with Intel Xeon Gold 6230 processors, 32GB of RAM, and 16 cores per node. The queueing system on Sapelo2 is Slurm.
  
−
For a subset of compute nodes, internodal communication among them and between these nodes and the storage systems serving the home directories and the scratch directories is provided by a QDR Infiniband network(40Gbps). For another subset of compute nodes, these communications are provided by an EDR Infiniband network.
+
Internodal communication among the compute nodes and between these nodes and the storage systems serving the home directories and the scratch directories is provided by an EDR Infiniband network (100Gbps).
  
  
Line 39: Line 39:
 
'''Regular nodes'''
 
'''Regular nodes'''
  
−
*106 compute nodes with AMD Opteron processors (48 cores and 128GB of RAM per node)  
+
* 119 compute nodes with AMD EPYC (Milan 3rd gen) processors (128 cores and 512GB of RAM per node)
−
* 22 compute nodes with AMD EPYC (Rome) processors (64 cores and 128GB of RAM per node)
+
* 4 compute nodes with AMD EPYC (Milan 3rd gen) processors (64 cores and 256GB of RAM per node)
−
* 16 compute nodes with AMD EPYC processors (32 cores and 128GB of RAM per node)
+
* 2 compute nodes with AMD EPYC (Milan 3rd gen) processors (64 cores and 128GB of RAM per node)
 +
* 123 compute nodes with AMD EPYC (Rome 2nd gen) processors (64 cores and 128GB of RAM per node)
 +
* 64 compute nodes with AMD EPYC (Naples 1st gen) processors (32 cores and 128GB of RAM per node)
 
* 42 compute nodes with Intel Xeon Skylake processors (32 cores and 192GB of RAM per node)
 
* 42 compute nodes with Intel Xeon Skylake processors (32 cores and 192GB of RAM per node)
−
* 32 compute nodes with Intel Xeon Broadwell processors (28 cores and 64GB of RAM per node)
+
 
−
*  4 compute nodes with AMD Opteron processors (48 cores and 256GB of RAM per node)
+
 
 +
'''High memory nodes (2TB/node)'''
 +
 
 +
*  2 compute nodes with AMD EPYC processors (32 cores and 2TB of RAM per node)
  
  
 
'''High memory nodes (1TB/node)'''
 
'''High memory nodes (1TB/node)'''
  
−
* 4 compute nodes with AMD EPYC processors (64 cores and 1TB of RAM per node)
+
* 2 compute nodes with AMD EPYC (Milan 3rd gen) processors (128 cores and 1TB of RAM per node)
−
* 4 compute nodes with Intel Xeon Broadwell processors (28 cores and 1TB of RAM per node)
+
* 5 compute nodes with AMD EPYC (Milan 3rd gen) processors (32 cores and 1TB of RAM per node)
−
* 1 compute node with AMD Opteron processors (48 cores and 1TB of RAM per node)
+
* 4 compute nodes with AMD EPYC (Naples 1st gen) processors (64 cores and 1TB of RAM per node)
 +
* 4 compute nodes with Intel Xeon Broadwell processors (28 cores and 1TB of RAM per node)
  
  
 
'''High memory nodes (512GB/node)'''
 
'''High memory nodes (512GB/node)'''
  
−
* 16 compute nodes with AMD EPYC processors (32 cores and 512GB of RAM per node)
+
* 18 compute nodes with AMD EPYC (Naples 1st gen) processors (32 cores and 512GB of RAM per node)
−
*  6 compute nodes with AMD Opteron processors (48 cores and 512GB of RAM per node)
 
 
<!-- *  1 compute node with Intel Xeon Nehalem processors (32 cores and 512GB of RAM per node) -->
 
<!-- *  1 compute node with Intel Xeon Nehalem processors (32 cores and 512GB of RAM per node) -->
  
Line 63: Line 68:
 
'''GPU nodes'''
 
'''GPU nodes'''
  
−
* 4 compute nodes with Intel Xeon Skylake processors (32 cores and 187GB of RAM) and 1 NVIDIA P100 GPU card per node
+
* 12 compute nodes with AMD EPYC (Milan 3rd gen) processors (64 cores and 1TB of RAM) and 4x NVIDIA A100 GPU cards.
−
* 2 compute nodes with Intel Xeon processors (16 cores and 128GB of RAM) and 8 NVIDIA K40m GPU cards per node  
+
* 2 compute nodes with Intel Xeon Skylake processors (32 cores and 187GB of RAM) and 1x NVIDIA P100 GPU card per node
−
* 4 compute nodes with Intel Xeon processors (12 cores and 96GB of RAM) and 7 NVIDIA K20Xm GPU cards per node  
+
<!-- * 2 compute nodes with Intel Xeon processors (16 cores and 128GB of RAM) and 8x NVIDIA K40m GPU cards per node -->
  
  
Line 72: Line 77:
 
* Various configurations
 
* Various configurations
  
−
 
−
The queueing system on Sapelo2 is Torque/Moab.
 
  
 
<!--
 
<!--
Line 95: Line 98:
 
====[[Disk Storage]]====
 
====[[Disk Storage]]====
  
−
====[[Software Installed on Sapelo2]]====
+
====[[Software on Sapelo2]]====
 +
 
 +
====[[Available Toolchains and Toolchain Compatibility]]====
  
 
====[[Code Compilation on Sapelo2]]====
 
====[[Code Compilation on Sapelo2]]====
Line 102: Line 107:
  
 
====[[Monitoring Jobs on Sapelo2]]====
 
====[[Monitoring Jobs on Sapelo2]]====
 +
 +
====[[Migrating from Torque to Slurm]]====
 +
 +
 +
 +
'''Training material'''
 +
 +
To help users familiarize with Slurm and the test cluster environment, we have prepared some training videos that are available from the GACRC's Kaltura channel at https://kaltura.uga.edu/channel/GACRC/176125031 (login with MyID and password is required). Training sessions and slides are available at https://wiki.gacrc.uga.edu/wiki/Training
  
  
Line 107: Line 120:
 
[[#top|Back to Top]]
 
[[#top|Back to Top]]
  
−
 
+
<!--
 
===  Slurm Test Cluster (Sap2test) ===
 
===  Slurm Test Cluster (Sap2test) ===
  
Line 136: Line 149:
 
'''Storage'''
 
'''Storage'''
  
−
The user's home directory (/home), scratch directory (/scratch), and each group's work directory (/work) on the Slurm test cluster are the same file systems as on Sapelo2. So there is no need to transfer data between Sapelo2 and Slurm test cluster. If you have Sapelo2 specific settings in your dotfiles (for example in .bashrc or in software specific configuration files), those might need to get changed when you work on Sap2test.
+
The user's home directory (/home), scratch directory (/scratch), and each group's work directory (/work) on the Slurm test cluster are the same file systems as on Sapelo2. So there is no need to transfer data between Sapelo2 and Slurm test cluster. If you have Sapelo2 specific settings in your dotfiles (for example in .bashrc or in software specific configuration files), those might need to get changed when you work on Sap2test. The environment variable GACRC_CLUSTER stores the test cluster name, and can be used to set up a cluster specific dotfile to use on the test cluster.
  
 
However, Sapelo2's /usr/local file system and therefore the applications installed on Sapelo2 are not available on the Slurm test cluster.  
 
However, Sapelo2's /usr/local file system and therefore the applications installed on Sapelo2 are not available on the Slurm test cluster.  
 +
 +
 +
'''Training material'''
 +
 +
To help users familiarize with Slurm and the test cluster environment, we have prepared some training videos that are available from the GACRC's Kaltura channel at https://kaltura.uga.edu/channel/GACRC/176125031 (login with MyID and password is required). Training sessions and slides are available at https://wiki.gacrc.uga.edu/wiki/Training
 +
 +
 +
'''Getting Help'''
 +
 +
If you run into any issues on the test cluster or have any questions or suggestions, please let me know via the online form below, as it will reach all the GACRC staff members:
 +
 +
[https://uga.teamdynamix.com/TDClient/2060/Portal/Requests/ServiceDet?ID=41600 Support for Slurm test cluster]
 +
  
 
====[[Connecting to the Slurm test cluster]]====
 
====[[Connecting to the Slurm test cluster]]====
Line 146: Line 172:
 
====[[Software on sap2test | Software Installed on the Slurm test cluster]]====
 
====[[Software on sap2test | Software Installed on the Slurm test cluster]]====
  
−
====[[Code Compilation on the Slurm test cluster]]====
+
====[[Code Compilation on Sap2test]]====
  
 
====[[Available Toolchains and Toolchain Compatibility]]====
 
====[[Available Toolchains and Toolchain Compatibility]]====
Line 159: Line 185:
  
  
−
If you run into any issues on the test cluster or have any questions or suggestions, please let me know via the online form below, as it will reach all the GACRC staff members:
 
−
 
−
[https://uga.teamdynamix.com/TDClient/2060/Portal/Requests/ServiceDet?ID=41600 Support for Slurm test cluster]
 
  
 
----
 
----
 
[[#top|Back to Top]]
 
[[#top|Back to Top]]
 +
 +
-->
  
 
===  Teaching cluster ===
 
===  Teaching cluster ===
  
−
The teaching cluster is a Linux cluster that runs a 64-bit Linux, with Centos 7.5. The physical login node has two 6-core Intel Xeon E5-2620 processors and 128GB of RAM and it runs Red Hat EL 7.5. An Ethernet network (1Gbps) provides internodal communication among compute nodes, and between the compute nodes and the storage systems serving the home directories and the work directories.
+
The teaching cluster is a Linux cluster that runs a 64-bit Linux, with Rocky 8.8. The login node is a VM that has 4 cores (Intel Xeon Gold 6230 processor) and 16GB of RAM. An EDR Infiniband network (100Gbps) provides internodal communication among compute nodes, and between the compute nodes and the storage systems serving the home directories and the work directories.
  
 
The cluster is currently comprised of the following resources:  
 
The cluster is currently comprised of the following resources:  
  
−
*37 compute nodes with Intel Xeon X5650 2.67GHz processors (12 cores and 48GB of RAM per node)  
+
'''Regular nodes:'''
−
* 2 compute nodes with Intel Xeon E5504 2.00GHz processors (8 cores and 48GB of RAM per node)
+
 
−
* 3 compute nodes with Intel Xeon E5504 2.00GHz processors (8 cores and 192GB of RAM per node)
+
* 10 compute nodes with AMD EPYC (Naples 1st gen) processors (32 cores and 128GB or RAM per node)
−
* 2 compute nodes with AMD Opteron 6174 processors (48 cores and 128GB of RAM per node)
+
 
−
* 3 compute nodes with AMD Opteron 6128 HE 2.00GHz processors (32 cores and 64GB of RAM per node)
+
'''High-memory nodes:'''
−
* 6 NVIDIA Tesla (Fermi) M2070 GPU cards (8 x 448 = 3584 GPU cores). These cards are installed on one host that has dual 6-core Intel Xeon CPUs and 48GB of RAM
+
 
 +
* 2 compute nodes with AMD EPYC (Naples 1st gen) processors (64 cores and 1TB of RAM per node)
 +
 
 +
'''GPU nodes:'''
 +
 
 +
* 1 compute node with Intel Skylake processors (32 cores, 192GB RAM per node) and a P100 GPU card
 +
<!--
 +
*30 compute nodes with Intel Xeon X5650 2.67GHz processors (12 cores and 48GB of RAM per node)  
 +
* 2 compute nodes with Intel Xeon L7555 1.87GHz processors (32 cores and 512GB of RAM per node)
 +
* 4 NVIDIA Tesla (Kepler) K20Xm GPU cards. These cards are installed on one host that has dual 6-core Intel Xeon CPUs and 48GB of RAM
 +
-->
  
 
The queueing system on the teaching cluster is Slurm.
 
The queueing system on the teaching cluster is Slurm.
Line 183: Line 218:
 
====[[Connecting#Connecting_to_the_teaching_cluster |Connecting to the teaching cluster]]====
 
====[[Connecting#Connecting_to_the_teaching_cluster |Connecting to the teaching cluster]]====
  
−
====[[Transferring Files]]====
 
  
 +
====[[Transferring Files]]==== 
 +
<!--
 
====[[Disk Storage]]====
 
====[[Disk Storage]]====
−
 
+
-->
 
====Software Installed on the teaching cluster====
 
====Software Installed on the teaching cluster====
  
−
The list of installed application is available at [[Software]] page.
+
The teaching cluster has access to the same software stack installed on Sapelo2.
  
 
====[[Code Compilation on the teaching cluster]]====
 
====[[Code Compilation on the teaching cluster]]====

Latest revision as of 10:44, 24 January 2024



Sapelo2

Sapelo2 is a Linux cluster that runs a 64-bit Rocky 8.8 operating system and it is managed using Warewulf. Several virtual login nodes are available, with Intel Xeon Gold 6230 processors, 32GB of RAM, and 16 cores per node. The queueing system on Sapelo2 is Slurm.

Internodal communication among the compute nodes and between these nodes and the storage systems serving the home directories and the scratch directories is provided by an EDR Infiniband network (100Gbps).


The cluster is currently comprised of the following resources:

Regular nodes

  • 119 compute nodes with AMD EPYC (Milan 3rd gen) processors (128 cores and 512GB of RAM per node)
  • 4 compute nodes with AMD EPYC (Milan 3rd gen) processors (64 cores and 256GB of RAM per node)
  • 2 compute nodes with AMD EPYC (Milan 3rd gen) processors (64 cores and 128GB of RAM per node)
  • 123 compute nodes with AMD EPYC (Rome 2nd gen) processors (64 cores and 128GB of RAM per node)
  • 64 compute nodes with AMD EPYC (Naples 1st gen) processors (32 cores and 128GB of RAM per node)
  • 42 compute nodes with Intel Xeon Skylake processors (32 cores and 192GB of RAM per node)


High memory nodes (2TB/node)

  • 2 compute nodes with AMD EPYC processors (32 cores and 2TB of RAM per node)


High memory nodes (1TB/node)

  • 2 compute nodes with AMD EPYC (Milan 3rd gen) processors (128 cores and 1TB of RAM per node)
  • 5 compute nodes with AMD EPYC (Milan 3rd gen) processors (32 cores and 1TB of RAM per node)
  • 4 compute nodes with AMD EPYC (Naples 1st gen) processors (64 cores and 1TB of RAM per node)
  • 4 compute nodes with Intel Xeon Broadwell processors (28 cores and 1TB of RAM per node)


High memory nodes (512GB/node)

  • 18 compute nodes with AMD EPYC (Naples 1st gen) processors (32 cores and 512GB of RAM per node)


GPU nodes

  • 12 compute nodes with AMD EPYC (Milan 3rd gen) processors (64 cores and 1TB of RAM) and 4x NVIDIA A100 GPU cards.
  • 2 compute nodes with Intel Xeon Skylake processors (32 cores and 187GB of RAM) and 1x NVIDIA P100 GPU card per node


Buy-in nodes

  • Various configurations


Connecting to Sapelo2

Transferring Files

Disk Storage

Software on Sapelo2

Available Toolchains and Toolchain Compatibility

Code Compilation on Sapelo2

Running Jobs on Sapelo2

Monitoring Jobs on Sapelo2

Migrating from Torque to Slurm

Training material

To help users familiarize with Slurm and the test cluster environment, we have prepared some training videos that are available from the GACRC's Kaltura channel at https://kaltura.uga.edu/channel/GACRC/176125031 (login with MyID and password is required). Training sessions and slides are available at https://wiki.gacrc.uga.edu/wiki/Training


Back to Top


Teaching cluster

The teaching cluster is a Linux cluster that runs a 64-bit Linux, with Rocky 8.8. The login node is a VM that has 4 cores (Intel Xeon Gold 6230 processor) and 16GB of RAM. An EDR Infiniband network (100Gbps) provides internodal communication among compute nodes, and between the compute nodes and the storage systems serving the home directories and the work directories.

The cluster is currently comprised of the following resources:

Regular nodes:

  • 10 compute nodes with AMD EPYC (Naples 1st gen) processors (32 cores and 128GB or RAM per node)

High-memory nodes:

  • 2 compute nodes with AMD EPYC (Naples 1st gen) processors (64 cores and 1TB of RAM per node)

GPU nodes:

  • 1 compute node with Intel Skylake processors (32 cores, 192GB RAM per node) and a P100 GPU card

The queueing system on the teaching cluster is Slurm.

Connecting to the teaching cluster

Transferring Files

Software Installed on the teaching cluster

The teaching cluster has access to the same software stack installed on Sapelo2.

Code Compilation on the teaching cluster

Running Jobs on the teaching cluster

Monitoring Jobs on the teaching cluster

Retrieved from "https://wiki.gacrc.uga.edu/index.php?title=Systems&oldid=21702"