[
https://issues.apache.org/jira/browse/CLOUDSTACK-9741?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
mrunalini updated CLOUDSTACK-9741:
----------------------------------
Description:
==============
KDDI has reported an issue where a detached volume in VMware( I coud confirm
that the issue affects XenServer as well in lab repro) fails to create one of
the snapshots if it there are two scheduled snapshots at the same time for the
same volume. One of that goes to Allocated state.
In Hoopa we have introduced a new feature for allowing simultaneous snapshots
for volumes which are "associated with a VM".
Support has already communicated to KDDI that simultaneous snapshots on a
volume is not supported in the version they are using which is 4.3.0.2, or
rather it is a limitation. KDDI is concerned that this is not documented.
Before having further discussions with KDDI, I have reproduced the issue lab
with ACP 4.3.0.2 + XenServer. I could notice that any combination of
weekly,daily and hourly simultaneous snapshot creation on a detached volume
results one of the snapshots in Allocated state forever.
It might be possibly to have the same result by executing the manual snapshots
at the same time.I would like to know the following:
1. Is this a limitation in ACP? Please share the details.
2. Does it have any dependency on hypervisor?
3. My repro of simultaneous snapshots for an attached volume in a version
earlier to Hoopa did work, what is the improvement in Hoopa?
I have the repro setup,DEBUG as well as TRACE logs available.
http://10.112.6.37:8080/client/
ACP: admin/password
SSH /MYSQL: root/password
logs and DB: /root/kddi-snapshot-issue
was:When two simultaneous snapshots are craetes for a same volume, same vm,
one snapshot will created at a time and other will picked up in next snapshot
poll run
> Simultaneous snapshots for detached volume
> ------------------------------------------
>
> Key: CLOUDSTACK-9741
> URL: https://issues.apache.org/jira/browse/CLOUDSTACK-9741
> Project: CloudStack
> Issue Type: Bug
> Security Level: Public(Anyone can view this level - this is the
> default.)
> Reporter: mrunalini
>
> ==============
> KDDI has reported an issue where a detached volume in VMware( I coud confirm
> that the issue affects XenServer as well in lab repro) fails to create one of
> the snapshots if it there are two scheduled snapshots at the same time for
> the same volume. One of that goes to Allocated state.
> In Hoopa we have introduced a new feature for allowing simultaneous snapshots
> for volumes which are "associated with a VM".
> Support has already communicated to KDDI that simultaneous snapshots on a
> volume is not supported in the version they are using which is 4.3.0.2, or
> rather it is a limitation. KDDI is concerned that this is not documented.
> Before having further discussions with KDDI, I have reproduced the issue lab
> with ACP 4.3.0.2 + XenServer. I could notice that any combination of
> weekly,daily and hourly simultaneous snapshot creation on a detached volume
> results one of the snapshots in Allocated state forever.
> It might be possibly to have the same result by executing the manual
> snapshots at the same time.I would like to know the following:
> 1. Is this a limitation in ACP? Please share the details.
> 2. Does it have any dependency on hypervisor?
> 3. My repro of simultaneous snapshots for an attached volume in a version
> earlier to Hoopa did work, what is the improvement in Hoopa?
> I have the repro setup,DEBUG as well as TRACE logs available.
> http://10.112.6.37:8080/client/
> ACP: admin/password
> SSH /MYSQL: root/password
> logs and DB: /root/kddi-snapshot-issue
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
