Im using XSIBackup-DC 220.127.116.11 it is installed on VMware 6.7.0 Update 3 (Build 14320388)
When I start an backup out of cron, sometimes I can find folowing messages in error.log
2020-04-08T02:52:44 | Error code 678 at file esxi.c, line 678
Error description: could not create snapshot in GIT-Server, error: KO
When this occured than the VM has 31 snapshots created by XSIBackup-DC after the run.
These must be deleted manually.
What means error KO?
How can I solve this Problem?
©XSIBackup-DC (as well as all other versions of ©XSIBackup) request an snapshot to be taken to the ©ESXi system. They don't create the snapshots, they just wait for the snapshot to complete before proceeding with the backup. As the information returned by ©ESXi is that the snapshot did not complete ©XSIBackup-DC believes that and returns an error.
Even in the event that you receive an error, some backup might still be taking place, we can't conclude that from the information you provide, as it will in turn depend on so many facts. The simplest way to find out is to look into your backups, check them via the --check argument and eventually restore some of them. Please be aware that you might need to arrange the .vmx file to point to some previous snapshot or .vmdk file, as when some error like that is raised the backup process might not be able to re-arrange the disk paths automatically. In any case a working -flat.vmdk is all you need to restore a VM.
Now the matter is why is your VM resisting or failing to be snapshotted?. You will need to get into the guest OS and check the logs or event viewer to retrieve information about the error that was raised at the moment to take the snapshot. This is usually caused by some service not being quiesced properly (in case you chose to quiesce the guest OS, which you did not state so far).
What is quiescing?
Well it's a process by which the OS asks some service, usually a busy one, to write whatever it has pending in its I/O queue and halt for some time (usually less than a couple of seconds) while the snapshot file is being dettached and prepared to be written to. This ensures that all data written to disk by the service, usually a DB one, is consistent in the backup.
Why would some quiesced snapshot fail?
©VMWare Tools needs to be installed as well as some additional components that your database server might need in order to be in coordination with the snapshot process. You may not have VMWare Tools installed, they might be outdated, you may be lacking some additional components your guest needs to perform this operation correctly. Windows servers are usually harder to configure than Linux ones, the latter will usually work like a charm just as long as you have ©VMWare Tools installed.
What if I have some ancient DB server that can't be quiesced by ©VMWare Tools?
Just perform a warm backup. That will turn the VM off, create a snapshot that will be consistent (as the VM is off), then it will turn the VM on on top of the snapshot and perform the backup. In the worst of the cases your server will be down for one or two minutes.
Thank you for the detailed answere. I will check all mentioned Points.
By the way it happends on Windows Clients and Linux Clients and VMware Tools are installed but I will check
The Problem is solved.
I did a wrong cron configuration because I set the Minute Parameter to '*' when hour was reached every Minute a new XSIBackup was started creates an snapshot and than stopped because it found XSIBackup already running.
So thanks for the help