-
Notifications
You must be signed in to change notification settings - Fork 722
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Adding support for S3 Backup URLs in SQL 2022 #714
base: master
Are you sure you want to change the base?
Conversation
Great initiative! Improvement suggestion: support MAXTRANSFERSIZE up to 20 MB as supported by the S3 backend. Currently because of the parameter checks (max transfer size 4 MB) only databases up to 40 GB in size can be backed up. If I'm not mistaken. |
...if you don't provide multiple URLs, that is. https://learn.microsoft.com/en-us/sql/relational-databases/backup-restore/sql-server-backup-to-url-s3-compatible-object-storage?view=sql-server-ver16#part-numbers-and-file-size-limitations |
Great find! The new commit has checks specific to MAXTRANSERSIZE range for S3 as well as unsupported INIT option. |
I have modified the jobs on our test servers to backup straight to S3 using the changed backup procedure as listed in the PR. It works most of the time, however on a server with multiple databases in sizes up to 800GB, it started crashing the availability group and causing AG failovers. I am using the following parameters in the call. Note that I am generating the command sting so I can do uppercase on server name and AG name as we have naming inconsistency in our environment and S3 names are case sensitive:
Have anybody else experienced this on SQL 2022? I mean the AG failovers when using backups of big databases to S3. |
Hello, do you have any forecast to merge this change ? Thanks. |
But you may also need to add the provision to allow s3:// and no @credential clause when it comes to errors it throws when you don't supply a credential for Azure. S3 does not require a credential be specified. |
Thank you, this modification has been made. |
I only just found this after hacking around myself... #629 "The IDENTITY should always be 'S3 Access Key' when using the S3 connector." Given the above... may want a credential check as something like this (then its similar to Azure block blob):
Note its only Azure Page Blob that requires a credential in the backup string, Azure Block Blob and S3 backups don't, thus this bit of code could be reverted:
Back to:
|
Modified logic checks for @url to allow s3:// format URLs to support SQL 2022. Tested with EC2 SQL 2022 standard.