Description of problem:
I would like to request the following feature\functionality in RHOS:
rhos-log-collector, which is going to be a tool that will collect all needed logs and system state from all nodes in the cluster pending on the configuration.
in a distributed environment, tool will collect all logs from cinder, glance swift and all other components pending on their location.
we have a similar tool in rhev which collect all logs from the engine (manager) and all hypervisors.
I discussed this with Lon Hohberger/Alan Pevec. Since we are including pssh/pslurp in RHOS now (bug # 1004458), this bug could be trivially solved via a simple bash script that calls pssh with sosreport on a set of hosts defined in a text file.
I believe Lon or Alan was going to write that script to parse packstack answers file to get a list of hosts, or the user could manually generate. It could go into the new sos-openstack-modules package where we would be storing the openstack specific sos modules.
An example of what a simple script might look like is available here:
Speaking of which, I don't see a bugzilla component for sos-openstack-modules.
Bryn, you were going to drive owning that component correct?
*** Bug 1025852 has been marked as a duplicate of this bug. ***
Created attachment 827851 [details]
Split log collector into its own subpackage so it can have explicit RPM dep on pssh.
A new bug, 1054073, was opened regarding the message you get in the output of rhos-log-collector, which points you to find the collected logs in a directory where actually there are only the local node logs.
Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.
For information on the advisory, and where to find the updated
files, follow the link below.
If the solution does not work for you, open a new bug report.