Having recently done our first full backup of our Enterprise 10.6.1 environment, yesterday I was surprised to find that the server had maxed out its C drive space. The culprit was Enterprise backup files in this portal folder:
The files in here just keep being created and created until they take up huge amounts of space - clearing it freed up 16gb. I gather from talking with local technical support that these are related to incremental backups, which we are not yet doing, but it seems an oversight to me to have Enterprise/portal filling up your critical server's C drive. I would suggest that there needs to be a solution found to this. In my view it should not rely on customers finding this problem when the C drive fills up, then having to manually clear several gb of files.
Of course backups are great, and it's great that we're able to do a full or incremental backup and restore our server/portal in case of disaster. I just think this needs looking at.
Once you run the DR tool, the expectation was that you'll be scheduling backups on a regular basis. In order to clear out those logs, you'll need to run the DR tool again to create a new backup, which deletes them:
There is an enhancement to improve this behavior:
ENH-000123248 Allow users to completely turn off the incremental backup at their discretion
Thanks for the response Jonathan.
This expectation/behaviour should be clearly indicated with an option/requirement to save incremental files on a storage drive or shared folder (as the backups themselves are), when you first run a backup and in order to front-foot the potential storage issue. I did not see any other warning, so it is left to the "I have a problem" stage to find this out. I don't think the C drive is always the best place for these files. No doubt however that they will make incremental backups easier, if that's the route one chooses, so that's a good thing. Our backups, once I have the opportunity to look at scheduling them (1 FTE for 2 organisations), are likely to be about monthly as our configs won't change very quickly once fully set up, and that's a lot of files to stack up in that folder on a server C drive that has in our case 100gb total space.
I should add this is Windows Server, not Linux, but thanks for the link.
Those logs are transaction logs and integral to the internal database used within Portal. You typically won't want a database writing to a share for internal processes, so there won't be an option to store the transaction logs on a share or outside the local file system of the database.
We are working on separate options for the DR tool, one that expects you to run them on a scheduled basis so the transaction logs won't be a problem, (replicating to a standby environment), while the other creates a backup for point-in-time restores, (backup/restore workflow). The latter would not cause the problem with the transaction logs.
The same information applies to Windows: