caviness:resources [UDCECC Wiki]

This page is read only. You can view the source, but not change it. Ask your administrator if you think this is wrong.
====== Resources ======

The design of Caviness is similar to previous community clusters.

===== Compute =====

Designed as a multi-generational system, over time Caviness will host a variety of nodes, differing not just in memory sizes and presence of coprocessors but also in processor microarchitecture.

==== Generation 1 ====

The baseline node specification comprises:

|**CPU**|(2) Intel Xeon E5-2695v4|
|**Cores**|18 per CPU, 36 per node|
|**Clock rate**|2.10 GHz|
|**CPU cache**|32 KiB L1 data and instruction caches per core; 256 KiB L2 cache per core; 45 MiB L3 cache|
|**Local storage**|910 GB ''/tmp'' partition on a 960 GB SSD|
|**Network**|(1) 1 Gbps ethernet port; (1) 100 Gbps Intel Omni-path port|

Three RAM sizes are present:

^ ^Qty^Specification^
|**RAM**|64|128 GiB (8 x 16 GiB) DDR4, 2400 MHz|
|**RAM**|55|256 GiB (8 x 32 GiB) DDR4, 2400 MHz|
|**RAM**|7|512 GiB (16 x 32 GiB) DDR4, 2133 MHz|

Some nodes include GPU coprocessors:

^ ^Qty^Specification^
|**Coprocessor**|10|(2) nVidia P100 GPGPU, 12 GiB, PCIe|

Additionally, two nodes with additional local storage are present to facilitate user testing of enhanced local storage media:

^ ^Qty^Specification^
|**Enhanced storage**|2|6.1 GB ''/nvme'' partition, RAID0 across (2) 3.2 TB Micron 9200 NVMe|

===== Networking =====

There are two private ethernet networks in the cluster.  A dedicated 1 Gbps network carries management traffic (remote power control of nodes, console access, etc.).  A dedicated 10 Gbps network carries all data traffic (NFS, job scheduling, SSH access) to the nodes.

A 100 Gbps Intel Omni-path network also connects all nodes.  The OPA network carries Lustre filesystem traffic as well as most MPI internode communications.  The network uses a fat tree topology employing six spine switches.  Each leaf switch (two per rack) features 12 leaf-to-spine uplink ports and 36 host ports (3:1 oversubscription).

===== Storage =====

Each rack of compute equipment added to Caviness is designed to add storage capacity to the cluster:

  * Lustre Object Storage Targets (OSTs) and Servers (OSSs)
  * NFS server

The addition of OSTs/OSSs increases the aggregate capacity and bandwidth of the ''/lustre/scratch'' filesystem.  Individual NFS servers provide distinct capacity and bandwidth but do not aggregate with existing capacity or bandwidth — in short, they're just "more space."

A discussion of each distinct kind of storage available to users is found below.  General usage scenarios for each include:

  * Home directory
    * User's personal software builds and installs
    * VALET package definitions for those software installs
  * Workgroup directory
    * Software builds/installs used by one or more members of a workgroup
    * VALET package definitions for those software installs
    * Members' data sets, job working directories, job summary directories
    * Project data sets, job working directories, job summary directories
  * Lustre scratch
    * Temporary storage of large data sets
    * Temporary storage of job working directories

See the [[software:start|software management strategies]] portion of this site for help setting-up a personal or workgroup software build/install directory.

==== Home directories ====

Each user is granted a home directory with a 20 GiB limit (quota).  Typically users will build software in their home directory.  The relatively low quota often means that users cannot (and should not) submit computational jobs from their home directories.  Home directories are mounted at the path ''/home/<uid_number>'', where ''<uid_number>'' is a user's Unix UID number (an integer value, use the ''id'' command to determine it).

<WRAP center round info 60%>
The Bash shell allows you to reference your home directory as ''~/'' in most commands.  For example, ''ls -al ~/'' displays a long listing of the all hidden and normally-visible files and directories inside your home directory.
</WRAP>

The home directory is the location of a few important files and directories:

^File^Description^
|''.bashrc''|Commands executed by any new Bash shell spawned for the user|
|''.bash_profile''|Commands executed specifically by a new login shell for the user|
|''.bash_history''|Saved sequence of commands the user has interactively entered at the shell prompt| 
|''.bash_udit''|Configuration file controlling UD-specific behaviors of the Bash shell|
|''.valet/''|Directory containing a user's personal VALET package definitions; does not exist by default, should be created by the user if wanted (automatically added to ''VALET_PATH'' by UD Bash login scripts)|
|''.zfs/snapshot/''|Directory containing historical //snapshots// of the home directory|

The use of [[abstract:zfs-snapshots|ZFS snapshots]] as backup copies of a home directory is discussed elsewhere.  In general, the editing of ''.bashrc'' and ''.bash_profile'' is discouraged, especially for the alteration of the ''PATH'' and ''LD_LIBRARY_PATH'' environment variables.

==== Workgroup directories ====

Each //workgroup// that purchases capacity in the cluster receives a //workgroup directory// with a quota in proportion to its level of investment in the cluster:  the more compute capacity purchased, the more space granted.  Workgroup directories are mounted at the path ''/work/<workgroup-id>'' on all nodes in the cluster.

<WRAP center round info 60%>
Once you've started a shell in a workgroup using the ''workgroup -g <workgroup-id>'' command the ''WORKDIR'' environment variable contains the path to the workgroup directory.  This allows you to reference it in commands like ''ls -l ${WORKDIR}/users''.

Adding the ''-c'' flag to the ''workgroup'' command automatically starts the workgroup shell in ''$WORKDIR''.
</WRAP>

The typical layout of a workgroup directory often includes:

^Subdirectory^Description^
|''/work/<workgroup-id>/.zfs/snapshot/''|Always present, contains historical //snapshots// of the the workgroup directory|
|''/work/<workgroup-id>/sw/''|A directory to hold software used by multiple members of the workgroup|
|''/work/<workgroup-id>/sw/valet/''|Directory for VALET package definitions (automatically added to ''VALET_PATH'' by UD Bash login scripts)|
|''/work/<workgroup-id>/users/''|Directory to contain per-user storage rather than having them exist directly under ''/work/<workgroup-id>''|
|''/work/<workgroup-id>/projects/''|Directory to contain per-project storage rather than having them exist directly under ''/work/<workgroup-id>''|

None of these directories are mandatory, but they do tend to make management of a workgroup's resources easier.  In particular, the fact that ''/work/<workgroup-id>/sw/valet'' will be automatically added to the VALET search path means workgroup users do not need to alter ''VALET_PATH'' manually, in their ''.bashrc''/''.bash_profile'', or in their job scripts.

The use of [[abstract:zfs-snapshots|ZFS snapshots]] as backup copies of a workgroup directory is discussed elsewhere.

==== Lustre scratch ====

The ''/lustre/scratch'' file system is a high-speed parallel file system accessible from all nodes in the cluster.  Users/groups are free to create their own top-level directories under ''/lustre/scratch'' and are responsible for:

  * setting appropriate permissions to restrict access accordingly
  * removing files/directories that are no longer needed to keep capacity available over time

The total capacity can be checked using the ''lfs df'' command:

<code>
$ lfs df
UUID                   1K-blocks        Used   Available Use% Mounted on
scratch-MDT0000_UUID  2989410560   977815040  2011593472  33% /lustre/scratch[MDT:0]
scratch-OST0000_UUID 51093819392 12543377408 38549933056  25% /lustre/scratch[OST:0]
scratch-OST0001_UUID 51093630976 11691081728 39402001408  23% /lustre/scratch[OST:1]
scratch-OST0002_UUID 51093489664 12198661120 38894347264  24% /lustre/scratch[OST:2]
scratch-OST0003_UUID 51093485568 12000202752 39092807680  23% /lustre/scratch[OST:3]

filesystem_summary:  204374425600 48433323008 155939089408  24% /lustre/scratch
</code>

Note that this command displays both aggregate capacity and the capacity of each OST and MDT (MetaData Target) component of the file system.  Users can determine their current occupied Lustre scratch capacity:

<code>
$ lfs quota -u $(id -u) /lustre/scratch
Disk quotas for usr 1001 (uid 1001):
     Filesystem  kbytes   quota   limit   grace   files   quota   limit   grace
/lustre/scratch  313298       0       0       -      78       0       0       -
</code>

Likewise, capacity associated explicitly with a workgroup can be checked:

<code>
$ lfs quota -g $(id -g) /lustre/scratch
Disk quotas for grp 1001 (gid 1001):
     Filesystem  kbytes   quota   limit   grace   files   quota   limit   grace
/lustre/scratch  267784       0       0       -      16       0       0       -
</code>

<WRAP center round important 60%>
UD IT staff reserve the right to perform emergency removal of data from ''/lustre/scratch'' if occupied capacity reaches unsafe levels.  Periodic automated cleanup policies may become necessary if such levels persist.
</WRAP>