Worker node name has some limitation

Bug #1820810 reported by zhipeng liu
10
This bug affects 1 person
Affects Status Importance Assigned to Milestone
StarlingX
Fix Released
Medium
zhipeng liu

Bug Description

Brief Description
-----------------
When I named worker node as worker-0, it failed when I unlock worker node.

Severity
--------
Major

Steps to Reproduce
------------------
According to deployment guide to deploy bare metal environment, set worker node name as worker-0
------------------

Actual Behavior
----------------
This name should be acceptable

Revision history for this message
zhipeng liu (zhipengs) wrote :

In sysinv.log,
We can find root cause
File "/usr/lib64/python2.7/site-packages/sysinv/common/ceph.py", line 559, in update_crushmap
    storage_num = int(hostupdate.ihost_orig['hostname'][8:])
ValueError: invalid literal for int() with base 10: ''

If we name node as compute-0 it can get storage_num = 0
If we name node as worker-0 it cannot get storage_num

We should change the code here to fix the bug.
sysinv/common/ceph.py", line 559

Revision history for this message
Ghada Khalil (gkhalil) wrote :

Marking as release gating; this should be addressed to allow any name to be used.

tags: added: stx.2019.05 stx.config
Changed in starlingx:
status: New → Triaged
importance: Undecided → Medium
assignee: nobody → Cindy Xie (xxie1)
Revision history for this message
Ghada Khalil (gkhalil) wrote :

Assigning to Cindy to identify a prime. This can go to anyone ramping on sysinv.

zhipeng liu (zhipengs)
Changed in starlingx:
assignee: Cindy Xie (xxie1) → zhipeng liu (zhipengs)
Revision history for this message
OpenStack Infra (hudson-openstack) wrote : Fix proposed to stx-config (master)

Fix proposed to branch: master
Review: https://review.openstack.org/649332

Changed in starlingx:
status: Triaged → In Progress
Ken Young (kenyis)
tags: added: stx.2.0
removed: stx.2019.05
Revision history for this message
OpenStack Infra (hudson-openstack) wrote : Fix merged to config (master)

Reviewed: https://review.opendev.org/649332
Committed: https://git.openstack.org/cgit/starlingx/config/commit/?id=7d6de50af98d1e08b8d438420eb43c609c39bd45
Submitter: Zuul
Branch: master

commit 7d6de50af98d1e08b8d438420eb43c609c39bd45
Author: zhipengl <email address hidden>
Date: Wed Apr 3 06:33:46 2019 +0800

    Fix unlock for worker node when name is not compute-<no>

    Unlock fails for worker node that do not follow the usual hostname
    naming scheme of compute-<number>. When using for e.g. worker-0 instead
    of compute-0, unlock fails.

    Error message is:
    File "/usr/lib64/python2.7/site-packages/sysinv/common/ceph.py",
    line 559, in update_crushmap
        storage_num = int(hostupdate.ihost_orig['hostname'][8:])
    ValueError: invalid literal for int() with base 10: ''

    Test pass with this fix for worker-0 unlock test.

    Change-Id: I01865d43f4456de81d72689f799441531a444be2
    Closes-Bug: #1820810
    Signed-off-by: zhipengl <email address hidden>

Changed in starlingx:
status: In Progress → Fix Released
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.