- Clustered Data ONTAP 8.3
- ONTAP 9
- ONTAP Select
- SnapMirror updates, both manual and scheduled, will retry when encountering an abort condition.
- The default number of retries is eight.
- When troubleshooting the number of retries, and the resulting messages, can increase the amount of time required to reproduce an error, or locate a root cause.
- The number of SnapMirror retries is configurable, and lowering the number of retries will reduce the time required to troubleshoot SnapMirror issues.
- Lowering the number of retries can be particularly helpful in scenarios where we need to quickly reach an error when packet traces are being collected.
- SnapMirror relationship are managed by their respective SnapMirror policies,
- Retries are controlled by these same policies.
In the example below, SnapMirror is failing with network errors, so we are performing packet traces. As packet traces can grow large very quickly, we ideally want to start the traces, reproduce the error, and the stop traces to keep their size manageable. This is an ideal situation to reduce retries.
network error(CSM: Request was canceled because of a connection failure.)
network error(CSM: An operation did not complete within the specified timeout window.))
CSM: Connection aborted.