cloudstack-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From anil lakineni <>
Subject Re: [PROPOSAL] Snapshot Improvements
Date Sun, 26 Jul 2015 10:02:56 GMT
Hi Anshul and All,

I knew it is off topic here but you guys have good knowledge on snapshots
in CloudStack. So that I pointed my issues in this thread, Please help me.

I would like to know necessary steps to be taken after getting below list
of errors while creating volume-snapshot,

In the CloudStack web UI,snapshots(scheduled recurring daily volume
snapshot policy)

a. Alert
I was removing entries from UI and updating in Cloud Database (if any
entries related to this snapshot) to destroy state.In this case,tried to
remove entry from the UI if not happened am hacking database and updating
this snapshot state to destroy (snapshot and snapshot_store_ref)

b. Allocated
I was removing entries from UI and updating in Cloud Database (if any
entries related to this snapshot) to destroy state. In this case, removing
snapshot from the UI throwing errors so that am updating in the database
(select state from snapshot where id=) to BackedUp then able to remove it
from UI with some exceptions and then after updating the database snapshot
table's state to destroy. Here no snapshot_store_ref table entries for this
c. backingup
I was removing entries from UI and Cloud Database (if any entries related
to this snapshot) to destroy state . In this case,entries are being created
in snapshot & snapshot_store_ref tables. SO that am updating on both tables
state's to destroy then deleting vhd file from secondary storage as it was
not fully completed snapshot vhd. entries regarding about snapshot info on UI and database
In this case, am doing cloud management service restart,there after next
scheduled volume snapshots are creating but not already skipped volume
snapshots for the day. Reason for this we could not identified till date,
Even in management logs also no information.

In all the above cases, the reason could be network disruptions or any
other things. we are still investigating on this (please let us know if you
are suspecting any reasons).

My question is on the above cases,

1. Is it possible to re-initiate snapshot creation/continue in any case of
above listed errors because these are scheduled snapshot policy (daily)

2. Was all the process taking on above errors to remove error snapshots,
correct? if not please suggest solutions

Please review and suggest solutions


On Fri, Jul 24, 2015 at 11:12 PM, Mike Tutkowski <> wrote:

> Is there currently any mechanism in place to prevent users from trying to
> take VM snapshots and volume snapshots at the same time? For example, is
> there a check in place in the relevant APIs that throws an exception if you
> try to get yourself in this situation?
> On Fri, Jul 24, 2015 at 1:54 AM, Anshul Gangwar <
> >
> wrote:
> > Rajani,
> >
> > Because there were observed issues which are leading to VM corruption if
> > both type of snapshots are allowed to exist together.
> > I will be fixing issues to make sure VM corruption do not happen.
> >
> > Regards,
> > Anshul
> >
> > On 24-Jul-2015, at 12:36 pm, Rajani Karuturi <<mailto:
> >>> wrote:
> >
> > Hi Anshul,
> > Do you know why volume and VM snapshots werent allowed to co-exists?
> > What is being done to resolve them?
> >
> > ~Rajani
> >
> > On Thu, Jul 23, 2015 at 2:57 PM, Anshul Gangwar <
> > <>>
> > wrote:
> >
> > I am working on improving snapshots experience in CloudStack.
> >
> > FS is available at
> >
> >
> > .
> >
> > Please review and provide comments/suggestions.
> >
> > Regards,
> > Anshul
> >
> >
> >
> >
> --
> *Mike Tutkowski*
> *Senior CloudStack Developer, SolidFire Inc.*
> e:
> o: 303.746.7302
> Advancing the way the world uses the cloud
> <>*™*

  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message