In this example, we show how to launch the suite u-cs809 in the ARE VDI session using the persistent session.
If you don't know how to start the ARE VDI session at the first place, please follow this page.
If you don't have any persistent sessions alive, follow the instructions here to create one. Again, if you already have a persistent session alive, you don't have to create a new one for running the suite u-cs809. It needs only the standard project membership: hr22 , ki32 , ki32_mosrs and access .
0. Launch a new terminal in the ARE VDI
Click the MATE Terminal icon to launch a terminal in which you can run the commands showing below.
1. Load Modules
First we login to the persistent session `demo.abc111.xy99.ps.gadi.nci.org.au` with X11 forwarding enabled.
[abc111@gadi-cpu-bdw-0005 ~] $ ssh -Y demo.abc111.xy99.ps.gadi.nci.org.au
Then load the modules in the shell running inside the persistent session.
[abc111@demo ~]$ module use /g/data/hr22/modulefiles/ [abc111@demo ~]$ module load cylc7/23.09 Loading cylc7/23.09 Loading requirement: mosrs-setup/1.0.1
2. Clean up old directories used for u-cs809
Every cylc-run subdirectory on $HOME, by default, is a symlink to the real cylc-run directory for the suite on /scratch. Make sure you delete the directory on /scratch and then the symlink on $HOME itself.
- If you have not run any Cylc 7 suites or Cylc 8 workflows at NCI, you will not have the
~/cylc-rundirectory at all. In this case, proceed to step 3. If you would like to run the suite for the second time with some modifications, run the following commands to clean up the previous runtime files and proceed to step 4.
[abc111@demo ~]$ cd ~/roses/u-cs809 [abc111@demo u-cs809]$ rose suite-clean Clean u-cs809? y/n (default n) y [INFO] delete: localhost:cylc-run/u-cs809/work [INFO] delete: localhost:cylc-run/u-cs809/share/cycle [INFO] delete: localhost:cylc-run/u-cs809/share [INFO] delete: localhost:cylc-run/u-cs809 [INFO] delete: localhost:/scratch/xy99/abc111/cylc-run/u-cs809
If you have some legacy directories from previous work which cannot be cleaned up using
rose suite-clean, do the following to make room for the new run.[abc111@gadi-cpu-bdw-0005 ~]$ ls -l ~/cylc-run/ lrwxrwxrwx 1 abc123 om02 35 Dec 24 16:22 u-cs809 -> /scratch/xy99/abc111/cylc-run/u-cs809 [abc111@gadi-cpu-bdw-0005 ~]$ rm -rf /scratch/xy99/abc111/cylc-run/u-cs809 [abc111@gadi-cpu-bdw-0005 ~]$ rm ~/cylc-run/u-cs809
3. Check out Suites
Before checkout a fresh copy of u-cs809, run `mosrs-auth` to provide your MOSRS credentials. If your cache is still valid, you may skip this step and checkout the suite directly. If you never used your MOSRS account on Gadi, the command `mosrs-auth` leads you through the configuration steps in which you need to enter your MOSRS username once and password twice.
$ mosrs-auth INFO: You need to enter your MOSRS credentials here so that GPG can cache your password. Please enter the MOSRS password for testuser: INFO: Checking your credentials using Subversion. Please wait. INFO: Successfully accessed Subversion with your credentials. INFO: Checking your credentials using rosie. Please wait. INFO: Successfully accessed rosie with your credentials.
If an old directory at ~/roses/u-cs809 exists, consider to remove it before checking out the suite. It's a good idea to start with a fresh copy in the test in order to avoid any accidental edits made to the old copy.
[abc111@demo roses]$ rm -rf ~/roses/u-cs809 [abc123@demo roses]$ rosie co u-cs809 [INFO] u-cs809: local copy created at /home/111/abc111/roses/u-cs809
4. Run Suites
It is now time to run the suite.
[abc123@demo roses]$ cd ~/roses/u-cs809 [abc123@demo u-cs809]$ rose suite-run [INFO] export CYLC_VERSION=7.9.7 [INFO] export ROSE_ORIG_HOST=demo.abc111.xy99.ps.gadi.nci.org.au [INFO] export ROSE_SITE=nci [INFO] export ROSE_VERSION=2019.01.7 [INFO] symlink: /scratch/xy99/abc111/cylc-run/u-cs809 <= /home/111/abc111/cylc-run/u-cs809 [INFO] create: log.20231009T235546Z [INFO] symlink: log.20231009T235546Z <= log [INFO] create: log/suite [INFO] create: log/rose-conf [INFO] symlink: rose-conf/20231009T105546-run.conf <= log/rose-suite-run.conf [INFO] symlink: rose-conf/20231009T105546-run.version <= log/rose-suite-run.version INFO] create: meta [INFO] install: meta [INFO] source: /home/111/abc111/roses/u-cs809/meta [INFO] install: rose-suite.info [INFO] source: /home/111/abc111/roses/u-cs809/rose-suite.info [INFO] create: app [INFO] install: app [INFO] source: /home/111/abc111/roses/u-cs809/app [INFO] install: suite.rc [INFO] REGISTERED u-cs809 -> /home/111/abc111/cylc-run/u-cs809 [INFO] create: share [INFO] create: share/cycle [INFO] create: work [INFO] chdir: log/ [INFO] ._. [INFO] | | The Cylc Suite Engine [7.9.7] [INFO] ._____._. ._| |_____. Copyright (C) 2008-2019 NIWA [INFO] | .___| | | | | .___| & British Crown (Met Office) & Contributors. [INFO] | !___| !_! | | !___. _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ [INFO] !_____!___. |_!_____! This program comes with ABSOLUTELY NO WARRANTY; [INFO] .___! | see `cylc warranty`. It is free software, you [INFO] !_____! are welcome to redistribute it under certain [INFO] [INFO] *** listening on http://demo.abc111.xy99.ps.gadi.nci.org.au:1990/ *** [INFO] [INFO] To view suite server program contact information: [INFO] $ cylc get-suite-contact u-cs809 [INFO] [INFO] Other ways to see if the suite is still running: [INFO] $ cylc scan -n 'u-cs809' demo.abc111.xy99.ps.gadi.nci.org.au [INFO] $ cylc ping -v --host=demo.abc111.xy99.ps.gadi.nci.org.au u-cs809 [INFO] $ ps -opid,args 2158739 # on demo.abc111.xy99.ps.gadi.nci.org.au [INFO] [abc111@demo u-cs809]
When the suite starts running, a GCylc window will pop up. If not, run cylc gui to launch the window and select "File→Open Another Suite" then "suite u-cs809" to show the suite progress.
5. Monitor the suite run
Right-click on any of the tasks to monitor the suite run.
After all jobs finish, simple checks, such as listing the exit status of all the PBS jobs, can be useful.
[abc111@gadi-cpu-bdw-0005 u-cs809]$ cd ~/cylc-run/u-cs809 [abc111@gadi-cpu-bdw-0005 u-cs809]$ grep "Exit Status" -R log/job log/job/1/recon/NN/job.out: Exit Status: 0 log/job/1/recon/01/job.out: Exit Status: 0 log/job/1/atmos/NN/job.out: Exit Status: 0 log/job/1/atmos/01/job.out: Exit Status: 0 log/job/1/fcm_make/NN/job.out: Exit Status: 0 log/job/1/fcm_make/01/job.out: Exit Status: 0
Note, PBS jobs can return with the zero "Exit Status" but still have errors depending on how the tasks are composed. Closer inspections are recommended to look into individual tasks.