starlingx
/
metal
Code Issues Proposed changes

Check system upgrade status by querying 'controller' instead of 'pxecontroller' 

Wipedisk is executed from services that communicate over management network
so, to avoid possible availability issues of PXE network, we change the interface
through which wipedisk checks if a system is upgrading by querying 'controller'
instead of 'pxecontroller'.

Change-Id: I86810fc612723353638e7938081e1ca80f261b13
Closes-Bug: 1860165
Signed-off-by: Ovidiu Poncea <ovidiu.poncea@windriver.com>
This commit is contained in:
Ovidiu Poncea 2020-06-22 18:22:41 +03:00
parent 240773cf5d
commit a62dfdd4ea
1 changed files with 1 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
mtce/src/scripts/wipedisk
View File

@ -74,7 +74,7 @@ WIPE_HDD="$pvs_to_delete $WIPE_HDD"
# journal partitions. Even if this script is also called during upgrades, there was no issue
# observed during that operation, so we skip wiping the journals during upgrades.
JOURNAL_DISKS=""
HOST_IN_UPGRADE=$(curl -sf http://pxecontroller:6385/v1/upgrade/$(hostname)/in_upgrade 2>/dev/null)
HOST_IN_UPGRADE=$(curl -sf http://controller:6385/v1/upgrade/$(hostname)/in_upgrade 2>/dev/null)
# The "ceph-disk list" command works even if the ceph cluster is not operational (for example if
# too many monitors are down) so we can grab journal info from the node, even in such scenarios.
# As a safety measure, we also wrap the command in a timeout command; it should never take long