You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository has been archived by the owner on Nov 9, 2020. It is now read-only.
Discovery was turned off in #1047 - see details there.
We need to bring it back for crash recovery scenario of managed plugin.
[ ] update to latest Docker client SDK - @pshahzeb
[ ] use client API with timeouts, if we do not get reply within it (say 2 secs) , consider we are in start up mode and skip volume discovery - @pshahzeb
[ ] fix an issue with using DriverName to match volumes. Driver name in managed plugin could be whatever, so we'd better use PropagateMount location (i.e. /mnt/vmdk)
[ ] bring recovery test back refcnt_test.sh:test_crash_recovery, and use "docker-runc kill" there
Note: Per moby/moby#31903 Docker engine API is fully functional only after the initialization is complete. Restore is in the middle of the initialization.. That's why we need to cut short all attempts to discover volumes if Docker engine does not reply in time
1. Using timout based context while using docker client apis on plugin side to
popluate the existing volume data
2. upgrading the docker client api version to 1.25
3. Removing environment variable to enable the volume discovery
Fixes#1050
1. Using timout based context while using docker client apis on plugin side to
popluate the existing volume data
2. upgrading the docker client api version to 1.25
3. Removing environment variable to enable the volume discovery
Fixes#1050
Discovery was turned off in #1047 - see details there.
We need to bring it back for crash recovery scenario of managed plugin.
[ ] update to latest Docker client SDK - @pshahzeb
[ ] use client API with timeouts, if we do not get reply within it (say 2 secs) , consider we are in start up mode and skip volume discovery - @pshahzeb
[ ] fix an issue with using DriverName to match volumes. Driver name in managed plugin could be whatever, so we'd better use PropagateMount location (i.e. /mnt/vmdk)
[ ] bring recovery test back refcnt_test.sh:test_crash_recovery, and use "docker-runc kill" there
// @pdhamdhere
The text was updated successfully, but these errors were encountered: