[Bugs] [Bug 1637743] New: Glusterd seems to be attempting to start the same brick process twice

bugzilla at redhat.com bugzilla at redhat.com
Tue Oct 9 22:35:49 UTC 2018


            Bug ID: 1637743
           Summary: Glusterd seems to be attempting to start the same
                    brick process twice
           Product: GlusterFS
           Version: mainline
         Component: glusterd
          Assignee: bugs at gluster.org
          Reporter: srangana at redhat.com
                CC: bugs at gluster.org

Description of problem:

A brick process log seems to indicate that it is processing the same CLI
arguments twice and hence starting the same brick twice although possibly
within different processes. This leads to the second instance of changelog
xlator failing to bind to its socket (or failure to initialize its RPC session)
and thus resulsting in xlator init failure. This further leads to
glusterfs_process_volfp failure and hence calling cleanup_and_exit.

When the above happens it causes several SEGV induced thread faults. As the
xlators are unloaded and symbols missing, things like gcov and backtrace cause
SEGVs when they fail to process address to symbol resolution.

One delta is that the CLI arguments for the second brick instance seems to have
a different brick port, but all other argument (including the log file) seem to
be the same.

It is entirely possible that there are 2 processes using the same log file, or
2 instances of the same brick being started up in the process (with different
ports), what is causing this needs further investigation to address the
resultant cores.

### Tests where this pattern of failure is noticed ###
  - Test:
  - Test: ./tests/bugs/glusterd/quorum-validation.t

You are receiving this mail because:
You are on the CC list for the bug.
You are the assignee for the bug.

More information about the Bugs mailing list