iSCSI connections are not reinitiated after host reboot
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
os-brick |
Fix Released
|
Medium
|
Sophie Huang |
Bug Description
In an iSCSI/multipath environment, a Cinder volume is attached to a tenant server (VM) on a compute host. After the compute host is rebooted, some of the connection(s) to the iSCSI portal(s) are not reinitiated, because of this, missing iSCSI device(s) are observed. This issue is observed for both stable Train and Ussuri releases.
When PS https:/
Exception in thread Thread-39:
Traceback (most recent call last):
File "/usr/lib/
self.run()
File "/var/lib/
super(Thread, self).run()
File "/usr/lib/
self.
File "/var/lib/
session, manual_scan = self._connect_
File "/var/lib/
'--op', 'new'))
File "/var/lib/
delay_
File "/var/lib/
result = self.__
File "/var/lib/
return execute_root(*cmd, **kwargs)
File "/var/lib/
return self.channel.
File "/var/lib/
raise exc_type(
oslo_concurrenc
Command: iscsiadm -m node -T iqn.xxxxxxxxxxxx -p xxx.xxx.
Exit code: 6
Stdout: ''
Stderr: 'iscsiadm: Could not make /etc/iscsi/
After the above PS is applied, the exception is handled, however the underline issue is still exist and missing iSCSI device(s) are observed.
Suggestion:
Is it possible to add some retries to _connect_
Changed in os-brick: | |
assignee: | nobody → Sophie Huang (sh879n) |
importance: | Undecided → Low |
importance: | Low → Medium |
Changed in os-brick: | |
status: | New → In Progress |
tags: | added: iscsi multipath reboot |
Please refer to this patchset for a potential fix: /review. opendev. org/c/openstack /os-brick/ +/810331
https:/