Watched the video on HA configuration. I plan to have a Primary Server run SquaredUp, with the Secondary server only used when the Primary Server is rebooted for patching, or if the Primary Server suffers an unscheduled outage.
I note that the HA File Share would be on a third server. In the video, SquaredUp talk about ensuring this is is either backed up, or has some HA itself.
It is said on the video: “Obviously, if you lose the file share completely, then nothing’s going to work, in which case you’d be, effectively, turning HA off and talking to your Storage Admin, frantically trying to recover a backup, because at this point you have lost all of your user content, your dashboards, anything you have made yourself”
Now, this suggests that SQUP would go down, if the HA file share was not available? But SQUP support said this last year:
"Usually when a file share in HA setup goes offline the data would be still available on your primary and secondary server disks and you can work on your primary server without interruption and when the share comes back online the data would be updated to the file share and the other secondary servers.
If for some reason the data gets wiped out, the SquaredUp servers can recreate the share at any moment with the available data. So, even if the share is not clustered you can always have the updated data."
So…confused here. Do I need to ensure my HA Fileshare is in two locations (prob via DFS) and replicate the fileshare data across these? To ensure the Fileshare is always available? Or will SquaredUp be absolutely fine with loss of the File Share for 10mins when the File Share server is patched and rebooted?
OK, I have had further confirmation from SquaredUp support, that their original statement was correct.
“HA simply handles syncing the two servers, so if it goes offline or the file is deleted you shouldn’t lose any data. In the case of deletion, the primary server will simply recreate the file with the contents of it’s user data. Though note any changes made on the secondary while the HA was down will be overwritten. You shouldn’t lose any user data or dashboards if there are issues with the file share.”
Actually you don’t lose any data but SQ will be offline during that time (and even after oftentimes). You will need to restart both primary and secondary in sequence (stop both, start primary, wait 10 minutes, boot secondary – as per support) as the SquaredUp DB will likely lock itself. Happens to me as I put the share on the primary node of the SCOM Always-On DB. When that one is patched, I lose access to SQ until I restart the app pools in the order listed above.
I’d suggest putting the HA files on a highly available share somewhere.
Interesting! At the moment I have pointed my HA to a DFS share. Once a day, I’m using DFS-R to replicate the data to the disabled partner in the DFS-N, so I at least have a copy on another box, if I lose the server hosting the share.
So far so good, but still yet to test what happens when the server hosting the DFS share is patched and rebooted…
True. Keep in mind that the data in the share shouldn’t change that much (unless a lot of people are doing dashboards all at the same time) so likely your regular file backup solution will suffice for backup purposes…