View source for BwUniCluster2.0/Hardware and Architecture

= Architecture of bwUniCluster 2.0 =

The bwUniCluster 2.0 is a parallel computer with distributed memory. Each node of system consists of at least two Intel Xeon processor, local memory, disks, network adapters and optionally accelerators (NVIDIA Tesla V100). All nodes are connected by a fast InfiniBand interconnect. In addition the file system
Lustre, that is connected by coupling the InfiniBand of the file server with the InfiniBand
switch of the compute cluster, is added to bwUniCluster 2.0 to provide a fast and scalable
parallel file system.

The operating system on each node is Red Hat Enterprise Linux (RHEL) 8.4. A number of additional software packages like e.g. SLURM have been installed on top. Some of these components are of special interest to end users and are briefly
discussed in this document. Others which are of greater importance to system
administrators will not be covered by this document.

The individual nodes of the system may act in different roles. According to the services supplied by the nodes, they are separated into disjoint groups. From an end users point of view the different groups of nodes are login nodes, compute nodes, file server nodes and administrative server nodes.

'''Login Nodes'''

The login nodes are the only nodes that are directly accessible by end users. These nodes
are used for interactive login, file management, program development and interactive pre-
and postprocessing. Two nodes are dedicated to this service but they are all accessible via
one address and a DNS round-robin alias distributes the login sessions to the
different login nodes.

'''Compute Node'''

The majority of nodes are compute nodes which are managed by a batch system. Users 
submit their jobs to the SLURM batch system and a job is executed when the required resources become available (depending on its fair-share priority).

'''File Server Nodes'''

The hardware of the parallel file system Lustre incorporates some file server nodes; the file
system Lustre is connected by coupling the InfiniBand of the file server with the independent InfiniBand switch of the compute cluster. In addition to shared file space there is also local storage on the disks of each node (for details see chapter "File Systems").

'''Administrative Server Nodes'''

Some other nodes are delivering additional services like resource management, external
network connection, administration etc. These nodes can be accessed directly by system administrators only.

= Components of bwUniCluster =

{| class="wikitable"
|- 
! style="width:9%"|
! style="width:13%"| Compute nodes "Thin"
! style="width:13%"| Compute nodes "HPC"
! style="width:13%"| Compute nodes "HPC Broadwell"
! style="width:13%"| Compute nodes "Fat"
! style="width:13%"| GPU x4
! style="width:13%"| GPU x8
! style="width:13%"| Login
|-
!scope="column"| Number of nodes
| 100 + 60
| 360
| 352
| 6
| 14
| 10
| 4
|-
!scope="column"| Processors
| Intel Xeon Gold 6230
| Intel Xeon Gold 6230
| Intel Xeon E5-2660 v4
| Intel Xeon Gold 6230
| Intel Xeon Gold 6230
| Intel Xeon Gold 6248
|-
!scope="column"| Number of sockets
| 2
| 2
| 2
| 4
| 2
| 2
| 2
|-
!scope="column"| Processor frequency (GHz)
| 2.1 Ghz
| 2.1 Ghz
| 2.0 GHz
| 2.1 Ghz
| 2.1 Ghz
| 2.5 Ghz
| 
|-
!scope="column"| Total number of cores
| 40
| 40
| 28
| 80
| 40
| 40
| 40 / 20 (Broadwell)
|-
!scope="column"| Main memory
| 96 GB / 192 GB
| 96 GB
| 128 GB
| 3 TB
| 384 GB
| 768 GB
| 384 GB / 128 GB (Broadwell)
|-
!scope="column"| Local SSD
| 960 GB SATA
| 960 GB SATA
| 480 GB SATA
| 4,8 TB NVMe
| 3,2 TB NVMe
| 6,4 TB NVMe
| 
|-
!scope="column"| Accelerators
| -
| -
| -
| -
| 4x NVIDIA Tesla V100
| 8x NVIDIA Tesla V100
| 
|-
!scope="column"| Interconnect
| IB HDR100 (blocking)
| IB HDR100
| IB FDR
| IB HDR
| IB HDR
| IB HDR
| IB HDR100 (blocking)
|}
Table 1: Properties of the nodes

= File Systems =

On bwUniCluster 2.0 the parallel file system Lustre is used for most globally visible user data. Lustre is open source and Lustre solutions and support are available from different vendors. Nowadays, most of the biggest HPC systems are using Lustre. An initial home directory on a Lustre file system is created automatically after account activation, and the environment variable $HOME holds its name. Users can create so-called workspaces on another Lustre file system for non-permanent data with temporary lifetime. There is another workspace file system based on flash storage for special requirements available.

Within a batch job further file systems are available: 
* The directory $TMP is only available and visible on the local node. It is located on fast SSD storage devices.
* On request a parallel on-demand (BeeOND) file system is created which uses the SSDs of the nodes which were allocated to the batch job.
* On request the external LSDF Online Storage is mounted on the nodes which were allocated to the batch job. This file system is based on the parallel file system Spectrum Scale. 

Some of the characteristics of the file systems are shown in Table 2.

{| style="width:100%; vertical-align:top; background:#f5fffa;border:1px solid #000000;padding:1px"
|- style="width:20%;height=20px; text-align:left;padding:3px"
! style="background-color:#AAA;padding:3px"| Property
! style="background-color:#AAA;padding:3px"| $TMP
! style="background-color:#AAA;padding:3px"| BeeOND
! style="background-color:#AAA;padding:3px"| $HOME
! style="background-color:#AAA;padding:3px"| Workspace
! style="background-color:#AAA;padding:3px"| Workspace <br> on flash
|- 
| style="background-color:#d3ddd8;height=20px; text-align:left;padding:3px"| Visibility 
| style="height=20px; text-align:left;padding:3px"| local node
| style="height=20px; text-align:left;padding:3px"| nodes of batch job
| style="height=20px; text-align:left;padding:3px"| global
| style="height=20px; text-align:left;padding:3px"| global
| style="height=20px; text-align:left;padding:3px"| global
|- 
| style="background-color:#d3ddd8;height=20px; text-align:left;padding:3px"| Lifetime 
| style="height=20px; text-align:left;padding:3px"| batch job runtime
| style="height=20px; text-align:left;padding:3px"| batch job runtime
| style="height=20px; text-align:left;padding:3px"| permanent
| style="height=20px; text-align:left;padding:3px"| max. 240 days
| style="height=20px; text-align:left;padding:3px"| max. 240 days
|- style="vertical-align:top;"
| style="background-color:#d3ddd8;height=20px; text-align:left;padding:3px"| Disk space 
| style="height=20px; text-align:left;padding:3px"| 960 GB - 6.4 TB <br> details see table 1
| style="height=20px; text-align:left;padding:3px"| n*250 GB
| style="height=20px; text-align:left;padding:3px"| 1.2 PiB
| style="height=20px; text-align:left;padding:3px"| 4.1 PiB
| style="height=20px; text-align:left;padding:3px"| 236 TiB
|- 
| style="background-color:#d3ddd8;height=20px; text-align:left;padding:3px"| Capacity Quotas
| style="height=20px; text-align:left;padding:3px"| no
| style="height=20px; text-align:left;padding:3px"| no
| style="height=20px; text-align:left;padding:3px"| yes <br> 1 TiB per user <br> also per organization
| style="height=20px; text-align:left;padding:3px"| yes <br> 40 TiB per user
| style="height=20px; text-align:left;padding:3px"| yes <br> 5 TiB per user
|- 
| style="background-color:#d3ddd8;height=20px; text-align:left;padding:3px"| Inode Quotas
| style="height=20px; text-align:left;padding:3px"| no
| style="height=20px; text-align:left;padding:3px"| no
| style="height=20px; text-align:left;padding:3px"| yes <br> 10 million per user
| style="height=20px; text-align:left;padding:3px"| yes <br> 30 million per user
| style="height=20px; text-align:left;padding:3px"| yes <br> 25 million per user
|- 
| style="background-color:#d3ddd8;height=20px; text-align:left;padding:3px"| Backup
| style="height=20px; text-align:left;padding:3px"| no
| style="height=20px; text-align:left;padding:3px"| no
| style="height=20px; text-align:left;padding:3px"| yes
| style="height=20px; text-align:left;padding:3px"| no
| style="height=20px; text-align:left;padding:3px"| no
|- style="vertical-align:top;"
| style="background-color:#d3ddd8;height=20px; text-align:left;padding:3px"| Read perf./node
| style="height=20px; text-align:left;padding:3px"| 500 MB/s - 6 GB/s <br> depends on type of local SSD / job queue: <br> 520 MB/s @ single / multiple <br> 800 MB/s @ multiple_e <br> 6600 MB/s @ fat <br> 6500 MB/s @ gpu_4 <br> 6500 MB/s @ gpu_8
| style="height=20px; text-align:left;padding:3px"| 400 MB/s - 500 MB/s<br> depends on type of local SSDs / job queue: <br> 500 MB/s @ multiple <br> 400 MB/s @ multiple_e
| style="height=20px; text-align:left;padding:3px"| 1 GB/s
| style="height=20px; text-align:left;padding:3px"| 1 GB/s
| style="height=20px; text-align:left;padding:3px"| 1 GB/s
|- style="vertical-align:top;"
| style="background-color:#d3ddd8;height=20px; text-align:left;padding:3px"| Write perf./node
| style="height=20px; text-align:left;padding:3px"| 500 MB/s - 4 GB/s <br> depends on type of local SSD / job queue: <br> 500 MB/s @ single / multiple <br> 650 MB/s @ multiple_e <br> 2900 MB/s @ fat <br> 2090 MB/s @ gpu_4 <br> 4060 MB/s @ gpu_8
| style="height=20px; text-align:left;padding:3px"| 250 MB/s - 350 MB/s <br> depends on type of local SSDs / job queue: <br> 350 MB/s @ multiple <br> 250 MB/s @ multiple_e
| style="height=20px; text-align:left;padding:3px"| 1 GB/s
| style="height=20px; text-align:left;padding:3px"| 1 GB/s
| style="height=20px; text-align:left;padding:3px"| 1 GB/s
|- style="vertical-align:top;"
| style="background-color:#d3ddd8;height=20px; text-align:left;padding:3px"| Total read perf.
| style="height=20px; text-align:left;padding:3px"| n*500-6000 MB/s
| style="height=20px; text-align:left;padding:3px"| n*400-500 MB/s
| style="height=20px; text-align:left;padding:3px"| 18 GB/s
| style="height=20px; text-align:left;padding:3px"| 54 GB/s
| style="height=20px; text-align:left;padding:3px"| 45 GB/s
|- style="vertical-align:top;"
| style="background-color:#d3ddd8;height=20px; text-align:left;padding:3px"| Total write perf.
| style="height=20px; text-align:left;padding:3px"| n*500-4000 MB/s
| style="height=20px; text-align:left;padding:3px"| n*250-350 MB/s
| style="height=20px; text-align:left;padding:3px"| 18 GB/s
| style="height=20px; text-align:left;padding:3px"| 54 GB/s
| style="height=20px; text-align:left;padding:3px"| 38 GB/s
|}
---------------------------------------------------------------------------------------------------------
  global: all nodes of uc1 access the same file system;
  local: each node has its own file system;
  permanent: files are stored permanently;
  batch job: files are removed at end of the batch job.
---------------------------------------------------------------------------------------------------------
Table 2: Properties of the file systems

== Selecting the appropriate file system ==

In general, you should separate your data and store it on the appropriate file system.
Permanently needed data like software or important results should be stored below $HOME
but capacity restrictions (quotas) apply. In case you accidentally deleted data on $HOME
you can usually restore it from backup. Permanent data which is not needed for months
or exceeds the capacity restrictions should be sent to the LSDF Online Storage 
or to the archive and deleted from the file systems. Temporary data which is only needed on a single
node and which does not exceed the disk space shown in the table above should be stored
below $TMP. Data which is read many times on a single node, e.g. if you are doing AI training, 
should be copied to $TMP and read from there. Temporary data which is used from many nodes 
of your batch job and which is only needed during job runtime should be stored on a 
parallel on-demand file system. Temporary data which can be recomputed or which is the 
result of one job and input for another job should be stored in workspaces. The lifetime 
of data in workspaces is limited and depends on the lifetime of the workspace which can be 
several months.

For further details please check the chapters below.

== $HOME ==

The home directories of bwUniCluster 2.0 (uc2) users are located in the parallel file system Lustre.
You have access to your home directory from all nodes of uc2. A regular backup of these directories 
to tape archive is done automatically. The directory $HOME is used to hold those files that are
permanently used like source codes, configuration files, executable programs etc. 

On uc2 there is a default user quota limit of 1 TiB and 10 million inodes (files and directories) per user. 
You can chek your current usage and limits with the command
<pre>
$ lfs quota -uh $(whoami) $HOME
</pre>
In addition to the user limit there is a limit of your organization (e.g. university) which depends on the financial share. This limit is enforced with so-called Lustre project quotas. You can show the current usage and limits of your organization with the following command:
<pre>
lfs quota -ph $(grep $(echo $HOME | sed -e "s|/[^/]*/[^/]*$||") /pfs/data5/project_ids.txt | cut -f 1 -d\ ) $HOME
</pre>

== Workspaces ==

On uc2 workspaces can be used to store large non-permanent data sets, e.g. restart files or output
data that has to be post-processed. The file system used for workspaces is also the parallel file system Lustre. This file system is especially designed for parallel access and for a high throughput to large
files. It is able to provide high data transfer rates of up to 54 GB/s write and read performance when data access is parallel. 

Workspaces have a lifetime and the data on a workspace expires as a whole after a fixed period. The maximum lifetime of a workspace is 60 days, but it can be renewed at the end of that period 3 times to a total maximum of 240 days after workspace generation. If a workspace has inadvertently expired we can restore the data during a limited time (few weeks). In this case you should create a new workspace and report the name of the new and of the expired workspace in a ticket.

Creating, deleting, finding and extending workspaces is explained on the  [[workspace]] page.

On uc2 there is a default user quota limit of 40 TiB and 30 million inodes (files and directories) per user. 
You can chek your current usage and limits with the command
<pre>
$ lfs quota -uh $(whoami) /pfs/work7
</pre>
Note that the quotas include data and inodes for all of your workspaces and all of your expired workspaces (as long as they are not yet completely removed).

=== Reminder for workspace deletion ===

Normally you will get an email every day starting 7 days before a workspace expires. You can send yourself a calender entry which reminds you when a workspace will be automatically deleted:

 $ ws_send_ical.sh <workspace> <email>

== Improving Performance on $HOME and workspaces ==

The following recommendations might help to improve throughput and metadata
performance on Lustre filesystems.

=== '''Improving Throughput Performance''' ===

Depending on your application some adaptations might be necessary if you want to reach
the full bandwidth of the filesystems. Parallel filesystems typically stripe files over storage subsystems, i.e. large files are separated into stripes and distributed to different storage subsystems. In Lustre, the size of these stripes (sometimes also mentioned as chunks) is called stripe size and the number of used storage subsystems is called stripe count.

When you are designing your application you should consider that the performance of
parallel filesystems is generally better if data is transferred in large blocks and stored in
few large files. In more detail, to increase throughput performance of a parallel application
following aspects should be considered:

* collect large chunks of data and write them sequentially at once,

* to exploit complete filesystem bandwidth use several clients,

* avoid competitive file access by different tasks or use blocks with boundaries at stripe size (default is 1MB),

* if files are small enough for the SSDs and are only used by one process store them on $TMP.

With previous Lustre versions adapting the Lustre stripe count was the most important optimization. However, for the workspaces of uc2 the new Lustre feature Progressive File Layouts has been used to define file striping parameters. This means that the stripe count is adapted if the file size is growing. In normal cases users no longer need to adapt file striping parameters in case they have very huge files or in order to reach better performance. 

If you know what you are doing you can still change striping parameters, e.g. the stripe count, of a directory and of newly created files. New files and directories inherit the stripe count from the parent directory. E.g. if you want to enhance throughput on a single very large file which is created in the directory $HOME/my_output_dir you can use the command
<pre>
$ lfs setstripe -c-1 $HOME/my_output_dir
</pre>
to change the stripe count to -1 which means that all storage subsystems of the file system are used to store that file. If you change the stripe count of a directory the stripe count of existing files inside this
directory is not changed. If you want to change the stripe count of existing files, change
the stripe count of the parent directory, copy the files to new files, remove the old files
and move the new files back to the old name. In order to check the stripe setting of
the file my_file use
<pre>
$ lfs getstripe my_file
</pre>
Also note that changes on the striping parameters (e.g. stripe count) are not saved in the
backup, i.e. if directories have to be recreated this information is lost and the default stripe
count will be used. Therefore, you should annotate for which directories you made changes
to the striping parameters so that you can repeat these changes if required.

=== '''Improving Metadata Performance''' ===

Metadata performance on parallel file systems is usually not as good as with local
filesystems. In addition, it is usually not scalable, i.e. a limited resource. Therefore,
you should omit metadata operations whenever possible. For example, it is much better
to have few large files than lots of small files. In more detail, to increase metadata
performance of a parallel application following aspects should be considered:

* avoid creating many small files,

* avoid competitive directory access, e.g. by creating files in separate subdirectories for each task,

* if many small files are only used within a batch job and accessed by one process store them on $TMP,

* change the default colorization setting of the command ls (see below).

On modern Linux systems, the GNU ls command often uses colorization by default to
visually highlight the file type; this is especially true if the command is run within a terminal
session. This is because the default shell profile initializations usually contain an alias
directive similar to the following for the ls command:
<pre>
$ alias ls="ls --color=tty"
</pre>
However, running the ls command in this way for files on a Lustre file system requires
a stat() call to be used to determine the file type. This can result in a performance
overhead, because the stat() call always needs to determine the size of a file, and that
in turn means that the client node must query the object size of all the backing objects
that make up a file. As a result of the default colorization setting, running a simple
ls command on a Lustre file system often takes as much time as running the ls command
with the -l option (the same is true if the -F, -p, or the -classify option, or any other option 
that requires information from a stat() call, is used). To avoid this performance overhead
when using ls commands, add an alias directive similar to the following
to your shell startup script:
<pre>
$ alias ls="ls --color=never"
</pre>

== Workspaces on flash storage ==

There is another workspace file system for special reqirements available. The file system is called ''full flash pfs'' and is based on the parallel file system Lustre. Access will be granted on special request.

=== Advantages of this file system ===

# All storage devices are based on flash (no hard disks) with low access times. Hence performance is better compared to other parallel file systems for read and write access with small blocks and with small files, i.e. IOPS rates are improved.
# The file system is mounted on bwUniCluster 2.0 and HoreKa, i.e. it can be used to share data between these clusters.

=== Access restrictions ===

* Only HoreKa users or KIT users of bwUniCluster 2.0 can use this file system.
* Access is granted on request and for appropriate requirements. In order to request access, please open a ticket with the subject ''full flash pfs access required'' and describe the following topics:
** number of files you want to store on this file system
** needed capacity
** number of nodes used by your typical jobs
** special I/O requirements of your jobs

=== Using the file system ===

After access is granted, you can use the file system in the same way as a normal workspace. You just have to specify the name of the flash-based workspace file system using the option ''-F'' to all the commands that manage workspaces. On bwUniCluster 2.0 it is called ''ffuc'', on HoreKa it is ''ffhk''. For example, to create a workspace with name myws and a lifetime of 60 days on bwUniCluster 2.0 execute:
 ws_allocate -F ffuc myws 60

If you want to use the full flash pfs on bwUniCluster 2.0 '''and''' HoreKa at the same time, please note that you only have to manage a particular workspace on one of the clusters since the name of the workspace directory is different. However, the path to each workspace is visible and can be used on both clusters.

Other features are similar to normal workspaces. For example, we are able to restore expired workspaces for few weeks and you have to open a ticket to request the restore. There are quota limits with a default limit of 5 TB capacity and 25 millions inodes per user. You can check your current usage with
 lfs quota -uh $(whoami) /pfs/work8

== $TMP ==

While all tasks of a parallel application access the same $HOME and workspace directory, the 
$TMP directory is local to each node on bwUniCluster 2.0. All nodes have fast SSDs 
local storage devices which are used to store data below $TMP. Different tasks of a parallel
application use different $TMP directories when they do not utilize one node. This directory should
be used for temporary files being accessed by single tasks. It should also be used if you read the 
same data many times from a single node, e.g. if you are doing AI training. In this case you should 
copy the data at the beginning of your batch job to $TMP and read the data from there.
In addition, this directory should be used for the installation
of software packages. This means that the software package to be installed should be
unpacked, compiled and linked in a subdirectory of $TMP. The real installation of the
package (e.g. make install) should be made in(to) the Lustre filesystem. 

Each time a batch job is started, a subdirectory is created on each node and assigned to the job. 
$TMP is newly set and the name of the subdirectory contains the Job-id so that the
subdirectory name is unique for each job. This unique name is then assigned to the
environment variable $TMP within the job. At the end of the job the subdirectory is removed.
Although $TMP points to the same path name for different nodes of a job, the physical location 
on these nodes is different.

== LSDF Online Storage==

In some cases it is useful to have access to the LSDF Online Storage on the HPC-Clusters also. Therefore the LSDF Online Storage is mounted on the Login- and Datamover-Nodes.
Furthermore it can be used on the compute nodes during the job runtime with the constraint flag "LSDF" ([[bwUniCluster_2.0_Slurm_common_Features|Slurm common features]]
). There is also an example about the LSDF batch usage: [[bwUniCluster_2.0_Slurm_common_Features#LSDF_Online_Storage|Slurm LSDF example ]] .
<pre>
#!/bin/bash
#SBATCH ...
#SBATCH --constraint=LSDF
</pre>
<br>
For the usage of the LSDF Online Storage the following environment variables are available: $LSDF, $LSDFPROJECTS, $LSDFHOME.
Please request storage projects in the LSDF Online Storage seperately:
[https://www.lsdf.kit.edu/os/storagerequest/: LSDF Storage Request].

==BeeOND (BeeGFS On-Demand)==

Users of the HPC-Cluster have possibility to request a private BeeOND (BeeGFS) parallel filesystem for each job. The file system is created during job startup and purged after your job.

'''IMPORTANT: All data on the private filesystem will be deleted after your job. Make sure you have copied your data back to the global filesystem (within job), e.g., $HOME or any workspace.'''

BeeOND/BeeGFS can be used like any other parallel file system. Tools like cp or rsync can be used to copy data in and out. 

For detailed usage see here:[[BwUniCluster_2.0_Slurm_common_Features#BeeOND_.28BeeGFS_On-Demand.29|Request on-demand file system]]

==Backup and Archiving==

There are regular backups of all data of the home directories,whereas ACLs and extended attributes will
not be backuped. 

Please open a ticket if you need backuped data.


[[Category:bwUniCluster 2.0|File System]][[Category:Hardware_and_Architecture|bwUniCluster 2.0]]