Bug #9627
closedBug #9553: AssertionError "mon_thrash.py", line 143, in do_join" in upgrade:firefly-firefly-testing-basic-vps run
ceph_manager.py is missing
100%
Description
While trying to figure out 9553 (see below), I came across this problem:
CommandFailedError: Command failed on vpm155 with status 1: 'sudo ceph_objectstore_tool --data-path /var/lib/ceph/osd/ceph-1 --journal-path /var/lib/ceph/osd/ceph-1/journal --op list-pgs'
2014-09-29 19:22:50,088.088 WARNING:teuthology.run_tasks:Saw failure, going into interactive mode...
I ssh'ed to vpm155 and ran that command manually, and got:
sudo ceph_objectstore_tool --data-path /var/lib/ceph/osd/ceph-1 --journal-path /var/lib/ceph/osd/ceph-1/journal --op list-pgs
Traceback (most recent call last):
File "/usr/bin/ceph_objectstore_tool", line 9, in <module>
import ceph_managerceph_objectstore_tool is in ceph-qa-suite/tasks. ceph_manager.py is also in ceph-qa-suite/tasks
I betcha this ceph_manager.py is the ceph_manager that ceph_objectstore_tool wants to import.
Updated by Anonymous over 9 years ago
9553 is:
AssertionError "mon_thrash.py", line 143, in do_join" in upgrade:firefly-firefly-testing-basic-vps run
Updated by Anonymous over 9 years ago
- Status changed from New to 12
- Assignee set to David Zafman
Updated by Anonymous over 9 years ago
This might be due to things not getting automatically installed by chef on vps machines.
I will check this out today.
Updated by Anonymous over 9 years ago
I had an idea that possible teuthology changes that caused chef to not always run
may have caused this. After playing with this a few times, I do not believe that this is the case
Updated by David Zafman over 9 years ago
There is also a binary command ceph_objectstore_tool built from the ceph repo.
sudo ceph_objectstore_tool --data-path /var/lib/ceph/osd/ceph-1 --journal-path /var/lib/ceph/osd/ceph-1/journal --op list-pgs
This should be executing the binary command NOT ceph_objectstore_tool.py which is what appears to be happening because of the python error message.
Updated by David Zafman over 9 years ago
- Assignee changed from David Zafman to Anonymous
Updated by Anonymous over 9 years ago
- Status changed from 12 to In Progress
In progress. This may be a problem with previously improper suites. I need to check if this is still happening.
Updated by Anonymous over 9 years ago
- Status changed from In Progress to Duplicate
Okay. the original problem here has been solved. The problem that I am running into here appears to be the same problem happening in 9787. I will close this as a duplicate and update that other ticket.
Updated by Anonymous over 9 years ago
- Status changed from Duplicate to In Progress
Hmmm. 9787 was closed. i guess it makes sense to use this report to track things so I am reopening. David Zafman is making a tasks/ceph_manager.py change that is fixing this.
Updated by Anonymous over 9 years ago
- Status changed from In Progress to Resolved
The ceph_objectstore_tool issue existed in the runs prior to October 20, 2014. On October 20, 2014, the set ceph_objectstore_tool usage was set to false:
teuthology-2014-10-20_18:33:02-upgrade:dumpling-firefly-x:stress-split-giant-distro-basic-multi/561541/teuthology.log
On October 24, 2014 runs did not set ceph_objectstore_tool and did not report a ceph_objectstore_tool issue.
teuthology-2014-10-24_18:15:01-upgrade:dumpling-firefly-x:stress-split-giant-distro-basic-multi/569655/teuthology.log
Since then, the ceph_objectstore_tool issue has not cropped up.