Workspace migration

Note: This page describes the necessary steps to migrate workspaces to another workspace filesystems.

HPE Hawk will be finally shut down at the end of April and with it the availability of the workspace file system ws11 mounted on hawk and all the data located on it.

For projects that are to be continued from Hawk to the next generation supercomputer Hunter (HPE), it is therefore very important to also transfer the existing and required workspace data to Hunter (HPE).

User migration to new workspaces

On Hawk the workspace filesystem is ws11. On Hunter the workspace filesystem is ws12. The ws12 workspace filesystem from Hunter is also mounted on Hawk under the identical path names.

The first step is to create a workspace on the ws12 filesystem. Creation of workspaces on ws12 is only possible on the Hunter supercomputer. So, login to hunter, create a workspace and remenber the path name of your workspace by running the command ws_list -a. (see workspace mechanism)

The next steps are on Hawk. Now, your are able to migrate workspaces located on the ws11 filesystems onto the ws12 filesystem. Run the command ws_list -a on a hawk frontend system to display the path for all your workspaces.

Before you start

Migration for large amount of data consumes a lot of IO ressources. Please review and remove data not needed any more or move it into HPSS.

How to proceed / Time schedule for the gradual shutdown of ws10

• 2024-07-22 10:00:
  • The default workspace filesystem will be switched from ws10 to ws11. ws_* tools applied to ws10 require the additional option '-F <workspacefilesystem>'.
  • ws11 will get the same policy settings as ws10.
  • Quota limit settings for each user-group will be moved from ws10 to ws11 and the quota enforcement will be enabled on ws11! (Batchjobs will not be scheduled in case the quota was exceeded the limit for your group).
  • Max. extension for workspaces on ws10 will be reduced from 3 to 1 (ws_allocate, ws_extend).
  • Max. duration for workspaces on ws10 will be reduced from 60 days to 40 days (ws_allocate, ws_extend).

• 2024-08-19:
  • Max. durations for workspaces on ws10 will be reduced from 40 days to 10 days (ws_allocate).
  • Deactivation of ws_extend for ws10.

• 2024-09-09:
  • Max. duration for workspaces on ws10 will be reduced from 10 days to 3 days (ws_allocate).

• 2024-10-18:
  • Start removing all remaining data located on ws10.
  • Final shutdown.

Important remarks

• If you have to migrate data residing in workspaces from one to another filesystem, do not use the mv command to transfer data. For large amounts of data, this will fail due to time limits. Currently we recommend for e.g. millions of small files or for large amounts of data to use the following command inside a single node batch job: rsync -a --hard-links Old_ws/ new_ws/ .
• Please try to use the mpifileutils dcp or dsync.
• Take care when you create new batch jobs. If you have to migrate your workspace from an old filesystem to the new location, this takes time. Do not run any job while the migration process is active. This may result in inconsistent data.

Operation / Policies of the workspaces on ws11:

• No job of any user-group member will be scheduled for computation as long as the group quota is exceeded.
• Accounting.
• Max. lifetime of a workspace is currently 60 days.
• Default lifetime of a workspace is 1 day.
• Max. number of workspace extensions are 3.
• Please read related man pages or the online workspace mechanism document
  

in particular note that the workspace tools allow to explicitly address a specific workspace file system using the -F option (e.g. ws_allocate -F ws11.0 my_workspace 10)

• To list your available workspace file systems use ws_list -l.
• Users can restore expired workspaces using ws_restore.

Please read https://kb.hlrs.de/platforms/index.php/Storage_usage_policy

Using mpifileutils for data transfer

The mpifileutils suite provides MPI-based tools to handle typical jobs like copy, remove, and compare for large datasets, providing speedups of up to 50x compared to single process jobs. It can only be run on compute nodes via mpirun.

dcp or dsync is similar to cp -r or rsync; simply provide a source directory and destination and dcp / dsync will recursively copy the source directory to the destination in parallel.

dcp / dsync has a number of useful options; use dcp -h or dsync -h to see a description or use the [User Guide].

It should be invoked via mpirun.

We highly recommend to use dcp / dsync with an empty ~/.profile and ~/.bashrc only! Furthermore, take care that only the following modules are loaded when using mpifileutils (this can be achieved by logging into the system without modifying the list of modules and loading only the modules openmpi and mpifileutils):
1) system/site_names
2) system/ws/8b99237
3) system/wrappers/1.0
4) hlrs-software-stack/current
5) gcc/10.2.0
6) openmpi/4.1.4
7) mpifileutils/0.11

dcp

Parallel MPI application to recursively copy files and directories.

dcp is a file copy tool in the spirit of cp(1) that evenly distributes the work of scanning the directory tree, and copying file data across a large cluster without any centralized state. It is designed for copying files that are located on a distributed parallel file system, and it splits large file copies across multiple processes.

Run dcp with the -p option to preserve permissions and timestamps, and ownership.

-p  : preserve permissions and timestamps, and ownership

--chunksize C: Copy files larger than C Bytes in C Byte chunks (default ist 4MB)

We highly recommend using the -p option.

dsync

Parallel MPI application to synchronize two files or two directory trees.

dsync makes DEST match SRC, adding missing entries from DEST, and updating existing entries in DEST as necessary so that SRC and DEST have identical content, ownership, timestamps, and permissions.

--chunksize C: Copy files larger than C Bytes in C Byte chunks (default ist 4MB)

Job Script example

Here is an example of a job script.

You have to change the SOURCEDIR and TARGETDIR according to your setup. Also the number of nodes and wallclock time should be adjusted.

#!/bin/bash
#PBS -N parallel-copy
#PBS -l select=2:node_type=rome:mpiprocs=128
#PBS -l walltime=00:20:00

module load openmpi mpifileutils

SOURCEDIR=<YOUR SOURCE DIRECTORY HERE>
TARGETDIR=<YOUR TARGET DIRECTORY HERE>

sleep 5
nodes=$(cat $PBS_NODEFILE | sort -u | wc -l)
let cores=nodes*20

time_start=$(date "+%c  :: %s")
#mpirun -np $cores dcp -p --bufsize 8MB ${SOURCEDIR}/ ${TARGETDIR}/
mpirun -np $cores dsync --bufsize 8MB $SOURCEDIR $TARGETDIR
time_end=$(date "+%c  :: %s")

tt_start=$(echo $time_start | awk {'print $9'})
tt_end=$(echo $time_end | awk {'print $9'})
(( total_time=$tt_end-$tt_start ))
echo "Total runtime in seconds: $total_time"