i got 7 esx hosts and until now am using the xsibackup-free for my backups I was thinking to buy the pro version but end up buying the DC version, my mistake, i was expecting to able to use ti as the PRO backing up until 20 host. I realize that the license is valid only for 2 hosts. That's ok this was my misunderstanding.
XSIBackup-Dc is great I love the philosophy and practical use of it.
I need some advice on what is the best practice of a server/client setup.
To be more specific here are 2 working examples that from my perspective have the same result, Store backup the VM "vmhs-user" to "/vmfs/volumes/xsibkp145/esx89/".
so how the server/client feature is beneficial to my backup/restore workflow. What is the best practice of the server/client setup
"./xsibackup --backup "VMs(vmhs-user)" email@example.com:22:/vmfs/volumes/xsibkp145/esx89/ --firstname.lastname@example.org --subject="7.89-82" --use-smtp=1 --remote-path="/scratch/XSI/XSIBackup-DC/xsibackup"
"./xsibackup --backup "VMs(vmhs-user)" /vmfs/volumes/xsibkp145/esx89/ --email@example.com --subject="7.89-82" --use-smtp=1
Last edited by grhookgr (2020-10-30 19:07:36)
Well, it's true that you can achieve the same result using either of the two syntaxes: local or over IP.
Choosing what to use is left to you based on the knowledge of your set up. If you are in a LAN, you can attach your volume as an NFS datastore, nonetheless you will be adding some little latency to each I/O operation, which is not negligible.
On top of that, you will be using the same CPU to calculate hashes compress the data chunks and record them on the NFS share, while when you use (c)XSIBackup in a client/server mode you are balancing/ splitting the load on the CPU between the two servers.
As an extreme example, you could mount an NFS DS over a WAN, you won't get good results though, that's obvious. On the other side, using a client/server topology over a WAN will offer optimum results. What to do is up to you, in any case we believe choices are clear in every case.
Thanks for answering my question I really understand now how the server/client feature works.
one more question, is the --replica command a differential backup? (something like --onediff )
Last edited by grhookgr (2020-10-31 21:09:49)