ryan.miller at domaonline
Jul 31, 2013, 2:16 PM
Post #3 of 3
(42 views)
Permalink
|
I did remove all snap shots and ran the coalesce tool manually to no avail.
I've been fearing this would call for exporting the vm. I already exported the vm last night and plan to reimport tonight. How do I prevent this from happening in the future? I've googled over and over and haven't found much useful.
Ryan Miller
Systems Administrator
DOMA Technologies, LLC
Phone * 757.306.4920 x315
Cellular * 757.636.0033
Fax * 757.306.4922
www.domaonline.com<http://www.domaonline.com/>
From: Joseph Hom [mailto:jhom [at] softlayer]
Sent: Wednesday, July 31, 2013 3:52 PM
To: Ryan Miller; 'xen-api [at] lists'
Subject: RE: XCP 1.6 iSCSI storage issue
Do not delete the VDI. They are part of the snapshot chain and are needed. If you have snapshots, you may try to delete them. After some time the garbage collector may kick in and try to coalesce some of those vhds to free up space.
Another option, if you can afford the downtime, is to shutdown the vm, export it to another disk, delete the vm, and then import it. The export/import will collapse the vhd chain back to a single vhd.
From: xen-api-bounces [at] lists<mailto:xen-api-bounces [at] lists> [mailto:xen-api-bounces [at] lists] On Behalf Of Ryan Miller
Sent: Tuesday, July 30, 2013 7:40 AM
To: 'xen-api [at] lists'
Subject: [Xen-API] XCP 1.6 iSCSI storage issue
Good morning all,
I have a guest vm that is not able to take a snapshot as it is claiming insufficient space. The current configuration is as follows...
I created a 500gb volume on a Dell EQL.
232gb have been allocated for the windows 2008 vm, thus leaving more than half of the iSCSI volume for snapshots.
In XenCenter I see
[cid:image001.jpg [at] 01CE8E11]
I have another vm that has 75gb used of 100gb and it can take a snapshot without issue. I investigate where the 489.2gb is coming from and get the following output...
[root [at] localhos ~]# vhd-util scan -m "VHD-*" -f -c -l VG_XenStorage-6cbd33e6-e82
vhd=VHD-924dbfc3-9b48-4342-877a-16a033d4bfd1 capacity=249030508544 size=249095520256 hidden=1 parent=none
vhd=VHD-ab922300-b559-42f5-9193-66ed241cddd0 capacity=249030508544 size=4789895168 hidden=1 parent=VHD-924dbfc3-9b48-4342-877a-16a033d4bfd1
vhd=VHD-dc35b521-1400-428b-bbb7-5ce579eda40d capacity=249030508544 size=423624704 hidden=1 parent=VHD-ab922300-b559-42f5-9193-66ed241cddd0
vhd=VHD-656fd604-12b8-4951-ad68-b681acf1fd6f capacity=249030508544 size=19730006016 hidden=1 parent=VHD-dc35b521-1400-428b-bbb7-5ce579eda40d
vhd=VHD-cc9450ca-5edd-4f39-b632-839f75135bcb capacity=249030508544 size=1660944384 hidden=1 parent=VHD-656fd604-12b8-4951-ad68-b681acf1fd6f
vhd=VHD-69e01ab7-b117-43b2-8f44-27d85a9c5b29 capacity=249030508544 size=249523339264 hidden=0 parent=VHD-cc9450ca-5edd-4f39-b632-839f75135bcb
Sure enough the sum of these sizes adds up to 489gb.
Should I proceed by xe vdi-list and do xe vdi-forget for these VHDs? If so, is it safe to remove all the children VHDs and would this even solve our problem? Are these orphaned snapshots? Also, this may be a dumb question but when vdi-list displays read-only ( R0) : true, is it safe to say the guest vm is not using that vdi and delete it?
Any help or direction would be greatly appreciated. This vm happens to be our primary domain controller and the virtual full backup fails because of the snap shot error.
Ryan Miller
Systems Administrator
DOMA Technologies, LLC
Phone * 757.306.4920 x315
Cellular * 757.636.0033
Fax * 757.306.4922
www.domaonline.com<http://www.domaonline.com/>
|
image001.jpg
