Red Hat Bugzilla – Bug 1070629
Update `oo-accept-node` to point the adminstrator to the `oo-devel-node` set-quota command
Last modified: 2014-06-23 03:37:32 EDT
Description of problem: Update `oo-accept-node` to point the administrator to the `oo-devel-node` set-quota command Version-Release number of selected component (if applicable): How reproducible: if the quota is not set for any gear the `oo-accept-node` throws below error : FAIL: user 52f26244a61c379e26000001 does not have quotas imposed This should suggest the admins to run : oo-devel-node set-quota --with-container-uuid 52f26244a61c379e26000001 --blocks 1048576 --inodes 40000 Steps to Reproduce: 1. 2. 3. Actual results: oo-accept-node returns : FAIL: user 52f26244a61c379e26000001 does not have quotas imposed Expected results: This should suggest the admins to run : oo-devel-node set-quota --with-container-uuid 52f26244a61c379e26000001 --blocks 1048576 --inodes 40000 Additional info:
I just hate pointing users to oo-devel-node for administrative use. I wonder if we could add the needed functionality to oo-admin-ctl-gears or something. Also, obviously, gears shouldn't be created without quota (unless the deployment never set up quota to begin with) - is there a bug under which this happens?
I had an issue with my setup, where broker and node went out of sync. I had to reset `ntp`. After that I ran `oo-diagnostics -v` and at the step where 'oo-accept-node' runs, the issue occurred.
OK. I could definitely believe that happening due to an mcollective request getting dropped. In any case, this happened under normal operation, with perhaps adverse conditions. So we should tell people how to address it. oo-accept-node isn't big on giving repair advice, but I agree it should point you to whatever tool will help with this. oo-devel-node until we have a better option.
Should be addressed by https://github.com/openshift/origin-server/pull/5400
Commit pushed to master at https://github.com/openshift/origin-server https://github.com/openshift/origin-server/commit/6395a43075f06baf9c8e6df5cbe51ecb85969f78 oo-accept-node: Advise user re: missing user quota Bug 1070629 https://bugzilla.redhat.com/show_bug.cgi?id=1070629 Amend the missing user quota error message to advise admin on how to resolve
Ready for cherry-pick into enterprise
PR for enterprise-2.1 has been merged: https://github.com/openshift/enterprise-server/pull/264 https://github.com/openshift/enterprise-server/commit/f9affc2574af89187d30cac73331ac68de03744a
Verified on /2.1.z/2014-06-11.1 Steps to verify: 1.Create app; Get gear id is:uuid: 53993cbddb26c8df80000012 2.On the gear's node,delete the quota of the gear [root@node2 node]# edquota 53993cbddb26c8df80000012 3.run oo-accept-node Actual results: 2. Quota Deleted 3.Prompt the expected info: [root@node2 node]# oo-accept-node FAIL: user 53993cbddb26c8df80000012 does not have quotas imposed. This can be addressed by running: oo-devel-node set-quota --with-container-uuid 53993cbddb26c8df80000012 --blocks 1048576 --inodes 80000 1 ERRORS
Additioan info: After run the reccommended comand in step 3: oo-devel-node set-quota --with-container-uuid 53993cbddb26c8df80000012 --blocks 1048576 --inodes 80000 Run:#oo-accept-node Display:PASS
In the patch, it is using hard code to suggest user to set quota, is it possible to read quota setting value from /etc/openshift/resource_limits.conf to repalce the hard code value for set-quota command?
Johnny, Miheer has implemented the functionality you requested in comment 13 and backported it to enterprise-server/enterprise-2.1. PR: https://github.com/openshift/enterprise-server/pull/299 commit 82cabe049edaaa2bf61110d98fc0caed5a702b4b Author: Miheer Salunke <misalunk@redhat.com> Date: 2014-06-12 22:25:11 +0530 oo-accept-node: check_user: use configured quotas Update 'oo-accept-node' to point the administrator to the 'oo-devel-node' set-quota command using quota values fetched from resources_limit.conf. enterprise-server bug 1070629 origin-server #cherrypick origin-server commit fa50774c2b1ffb3f03f1df98dcbc23e42fdaee92
OK, according to comment 14, I move this bug to "MODIFIED" status, once package include the cherrypick is shipped in puddle, QE will verify it again.
http://etherpad.corp.redhat.com/puddle-2-1-2014-06-18
Verified on 2.1.z/2014-06-18.2 Steps to verify: 1.In order to confirm that quatos is really read from /etc/openshift/resource_limits.conf,could modify the quota_files=90000 or other number can be noticed. 2.Create an app; Get gear id:53a277db6892df929d000009 3.On the gear's node,delete the quota of the gear. #edquato 53a277db6892df929d000009 4.run oo-accept-node [root@node1 ~]# oo-accept-node FAIL: user 53a277db6892df929d000009 does not have quotas imposed. This can be addressed by running: oo-devel-node set-quota --with-container-uuid 53a277db6892df929d000009 --blocks 1048576 --inodes 90000 1 ERRORS (Notice that --inodes 90000 is indeed read from conf file.) 5.Run recommended command: [root@node1 ~]# oo-devel-node set-quota --with-container-uuid 53a277db6892df929d000009 --blocks 1048576 --inodes 90000 6.[root@node1 ~]# oo-accept-node PASS
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. http://rhn.redhat.com/errata/RHBA-2014-0781.html