2018-09-08:
The ccvm and ccvmcinder volumes on weizen-b are almost (but not quite) empty.
I don't know anything about the old CCVMs project, whether shrinking the volumes would invalidate part of that configuration, or if the data needs to be saved.
So I cowardly changed the space guarantee on the volumes from volume → none. Thus, they retain their original sizes, but the unused space is available to other volumes.
Copied volHvsc from pilsner-a → weizen-a. Updated mounts on mirror, www-node-{1,2}, and f5.
Started copy of volScene from pilsner-b → weizen-b.
2018-09-09:
Complete copy of volScene from pilsner-b → weizen-b. Updated mounts on mirror, www-node-{1,2}, and f5.
2018-09-27:
Started copy of volArchiveDebianOrg from pilsner-a → weizen-a. Disabled mirror updates.
2018-09-29:
Observed that the volArchiveDebianOrg copy finished. Updated mounts on mirror, www-node-{1,2}, and f5. Re-enabled mirror updates.
Remaining (non-mirror) volumes on pilsner-a:
Filesystem |
total |
used |
/vol/storage3_iscsi/ |
800GB |
776GB |
/vol/owncloud/ |
150GB |
0GB |
/vol/gitlab/ |
40GB |
0GB |
Details on the storage3_iscsi LUNs:
pilsner-a> lun show
/vol/storage3_iscsi/lun0 775.0g (832160399360) (r/w, online, mapped)
pilsner-a> igroup show
storage3 (iSCSI) (ostype: linux):
iqn.2009-08.org.cmucc:storage-3.initiator (not logged in)I guess we should set up an iSCSI client and give it access to see whether there's any data on the LUN.
Remaining (non-mirror) volumes on pilsner-b:
Filesystem |
total |
used |
/vol/retro_bt/ |
800GB |
791GB |
2019-02-10
Regrouping...
I did finally check out the storage3_iscsi LUN. It's empty/zeroes, so no problem deleting it.
So, it sounds like we want to:
- - See if anything exists in the owncloud and gitlab volumes. And back them up or move them to weizen. - See whether retro_bt matches retro_bt_backup already on weizen. If so, nothing to do there.
2019-02-16
There is some data in retro_bt that does not exist in retro_bt_backup.
So I've copied the owncloud and gitlab volumes to weizen-a, and am in the process of copying retro_bt to weizen-a.
The copy finished by the start of machine room hours, so we halted the pilsner controllers and powered the whole system off.
The disks have not been zeroed yet.