Closed DanielZhangQD closed 3 years ago
/assign @BinChenn
/unassign @BinChenn
/assign @csuzhangxc
@DanielZhangQD we remove backup data after archived (https://github.com/pingcap/tidb-operator/blob/6b44131af7eb94acb8cdaadf1e8847ba01200922/cmd/backup-manager/app/export/manager.go#L322) and remove archived data after uploaded to remote (https://github.com/pingcap/tidb-operator/blob/6b44131af7eb94acb8cdaadf1e8847ba01200922/cmd/backup-manager/app/export/manager.go#L350).
so, I think ad-hoc should not work now. Should we need to:
@csuzhangxc It's still useful with the current behavior. We have a user that has 16 TB backup, and it may take 48 hours to download the files from S3, and then the import with tidb-lightning failed, in this case, we can reuse the PVC, and no need to download the files again.
add an option to keep the backup data locally for ad-hoc
For this, we can support it if some users request this later, WDYT?
OK. I'll update docs for the case you mentioned above.
For this, we can support it if some users request this later, WDYT?
SGTM
Change Request
This repository is ONLY used to solve issues related to TiDB Operator documentation. For other issues (related to TiDB, PD, etc), please move to other repositories.
Please answer the following questions before submitting your issue. Thanks!
The data import with tidb-lightning chart supports the following three modes:
However, the ad-hoc mod is not covered in the current doc https://docs.pingcap.com/zh/tidb-in-kubernetes/stable/restore-data-using-tidb-lightning, which needs to update.
Describe your suggestion or addition.
Provide some reference materials (documents, websites, etc) if you could.