Point to these directories (this list from @agstephens):
Directory containing the sources from the GWS: /gws/smf/j04/cedaproc/cedawps/swallow/src
Directory containing the compiled executables and libraries: /gws/smf/j04/cedaproc/cedawps/swallow/build
Work directories for each job:
Test (on staging server): /gws/nopw/j04/ceda_wps/birds/test
Production (on cluster): /gws/nopw/j04/ceda_wps/birds/prod
Ensure that the directories under /gws/smf/j04/cedaproc/cedawps/swallow/ are accessible by users root and the service user (username in Slack message).
To add to the above, I am assuming that other files from the GWS (for example the MetRestore script) should go into other subdirectories of /gws/smf/j04/cedaproc/cedawps/swallow/ as appropriate.
Access control probably most easily set done via the top-level swallow directory. Need not be recursive.
Point to these directories (this list from @agstephens):
/gws/smf/j04/cedaproc/cedawps/swallow/src/gws/smf/j04/cedaproc/cedawps/swallow/build/gws/nopw/j04/ceda_wps/birds/test/gws/nopw/j04/ceda_wps/birds/prodEnsure that the directories under
/gws/smf/j04/cedaproc/cedawps/swallow/are accessible by usersrootand the service user (username in Slack message).To add to the above, I am assuming that other files from the GWS (for example the MetRestore script) should go into other subdirectories of
/gws/smf/j04/cedaproc/cedawps/swallow/as appropriate.Access control probably most easily set done via the top-level
swallowdirectory. Need not be recursive.