Today, the timeout value configurable on the sidecar applies to the regular attachment/detachment operations as well as the reconcile resyncs. However, with accounts that have a very large number of volumes in the backend, resyncs might take significantly longer to complete. We have such a case where, despite supporting paging at the maximum size possible, we needed to bump the resync timeout to multiple minutes to allow resyncs to complete at all. This conflicts with the timeout needs of attachments/detachments where a much lower value is appropriate.
Today, the timeout value configurable on the sidecar applies to the regular attachment/detachment operations as well as the reconcile resyncs. However, with accounts that have a very large number of volumes in the backend, resyncs might take significantly longer to complete. We have such a case where, despite supporting paging at the maximum size possible, we needed to bump the resync timeout to multiple minutes to allow resyncs to complete at all. This conflicts with the timeout needs of attachments/detachments where a much lower value is appropriate.