|
|
| Subject: | troubleshooting backup/restore problems by creating much smaller backups |
| Author: | Luke Lofgren |
| Posted: | 2018-06-25 11:37 |
I'm troubleshooting problems with "out of space" errors both backing up and restoring things. They all happen "at the end" of a large full backup (almost 1 TB) for both the restore and and the backup.
To speed things up I went ahead and used this command line to do a backup:
serviced backup --exclude hbase-* --exclude etl-analytics --exclude mariadb-events --exclude mariadb-model --exclude zencatalogservice /opt/serviced/var/backups
I am an enterprise customer, but I'm wanting to share more with the community. The above backup dramatically reduces what is being backed up (to the point its not really a backup anymore), but it allowed me to recreate the backup error quickly. Now I'm going to proceed to see if I can use it to recreate my error with a restore.
Version for the above: Control Center 1.5.1
| Subject: | RE: troubleshooting backup/restore problems by creating much smaller backups |
| Author: | Luke Lofgren |
| Posted: | 2018-06-25 12:12 |
| Subject: | RE: troubleshooting backup/restore problems by creating much smaller backups |
| Author: | Luke Lofgren |
| Posted: | 2018-06-25 13:55 |
This helped me self-diagnose that we were having problem with filling /var with docker images as part of backup/restore.
While a link for a "tmp" folder is standard install instructions, this actually seems better:
update /etc/sysconfig/docker to add a NEW line to the file which reads:
DOCKER_TMPDIR=/opt/serviced/var/backups/tmp
Also, doing backups that exclude hbase and other things make for more concise ways to do upgrade validation in non-PROD environments.
| < |
Previous Getting an error when select Devices under infrastructure menu |
Next How to get Dynamic View to work in Zenoss RM 4.2.5 Enterprise |
> |