[
https://issues.apache.org/jira/browse/CLOUDSTACK-5406?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13859226#comment-13859226
]
edison su commented on CLOUDSTACK-5406:
---------------------------------------
Hi Sanjay, seems you added this feature in 4.2:
https://cwiki.apache.org/confluence/display/CLOUDSTACK/Limit+Resources+to+domains+and+accounts,
could you help to fix it? The root cause is that we using volume.size() as the
snapshot size, which is incorrect, as the snapshot size usually is much smaller
than the volume size.
We can either disable the feature, or put the physical size of snapshot into
resource_count.
> Not able to take snapshot becasue of secondary_storage limit of 400 gb
> exceeded even though we have not really consumed this limit in secondary
> store.
> ------------------------------------------------------------------------------------------------------------------------------------------------------
>
> Key: CLOUDSTACK-5406
> URL: https://issues.apache.org/jira/browse/CLOUDSTACK-5406
> Project: CloudStack
> Issue Type: Bug
> Security Level: Public(Anyone can view this level - this is the
> default.)
> Components: Management Server
> Affects Versions: 4.3.0
> Environment: Build from 4.3
> Reporter: Sangeetha Hariharan
> Assignee: Sanjay Tripathi
> Priority: Critical
> Fix For: 4.3.0
>
> Attachments: storage.rar
>
>
> Set up:
> Xenserver setup with 2 NFS secondary stores.
> 1 account having 21 Vms.
> I am using the the default "secondary_storage" limit which is 400 gb.
>
> Started hourly snapshot policy for all the ROOT volumes of the Vms which was
> created from a template of size 20 gb. The actual used up size is 12 GB.
> When snapshot gets created , the full snapshot size is ~6.7 GB.
> After only 2 snapshots of ROOT volume , I hit the following excepton:
>
> “2013-12-06 00:00:48,298 WARN [c.c.s.s.SnapshotSchedulerImpl]
> (SnapshotPollTask:ctx-a7ccb50a) Scheduling snapshot failed due to
> com.cloud.exception.ResourceAllocationException: Maximum number of resources
> of type 'secondary_storage' for account
> name=test-TestParallelVolumeSnasohots-273V8U in domain id=1 has been
> exceeded.”
>
> How are we calculating this limit ? Do we query the actual secondary store /
> do we calculate based on the ROOT volume size and # of Snapshots.
>
> In case of Xenserve , we have Delta snapshots.We should not be
>
> mysql> select count(*) from snapshots where account_id=8 group by volume_id;
> +----------+
> | count(*) |
> +----------+
> | 2 |
> | 2 |
> | 2 |
> | 2 |
> | 2 |
> | 2 |
> | 2 |
> | 2 |
> | 2 |
> | 2 |
> | 2 |
> | 2 |
> | 2 |
> | 2 |
> | 2 |
> | 2 |
> | 2 |
> | 2 |
> | 2 |
> | 3 |
> | 3 |
> +----------+
> 21 rows in set (0.00 sec)
>
>
> [root@Rack3Host5 secondary]# du -h --max-depth=1
> 28G ./snapshots
> 13G ./template
> 4.0K ./volumes
> 41G .
>
>
> [root@Rack3Host8 secondary]# du -h --max-depth=1
> 12G ./template
> 29G ./snapshots
> 4.0K ./volumes
> 41G .
> [root@Rack3Host8 secondary]#
> mysql> select * from resource_limit;
> +----+-----------+------------+-------------------+---------------+
> | id | domain_id | account_id | type | max |
> +----+-----------+------------+-------------------+---------------+
> | 31 | NULL | 8 | user_vm | 200 |
> | 32 | NULL | 8 | public_ip | 200 |
> | 33 | NULL | 8 | volume | 200 |
> | 34 | NULL | 8 | snapshot | 200 |
> | 35 | NULL | 8 | template | 200 |
> | 36 | NULL | 8 | vpc | 200 |
> | 37 | NULL | 8 | cpu | 40 |
> | 38 | NULL | 8 | memory | 40960 |
> | 39 | NULL | 8 | network | 20 |
> | 40 | NULL | 8 | primary_storage | 214748364800 |
> | 41 | NULL | 8 | secondary_storage | 4294967296000 |
> +----+-----------+------------+-------------------+---------------+
> 11 rows in set (0.00 sec)
--
This message was sent by Atlassian JIRA
(v6.1.5#6160)
