What would you like to be added:
The flushAll method in the logpersister package should wait for all flushFromLastCheckpoint calls to complete before proceeding. Currently, it initiates these calls asynchronously using go sp.flushFromLastCheckpoint(ctx), which may lead to the cancel being executed immediately, not allowing the expected persister.gracePeriod(30-second timeout) to function correctly.
What would you like to be added: The
flushAll
method in thelogpersister
package should wait for allflushFromLastCheckpoint
calls to complete before proceeding. Currently, it initiates these calls asynchronously usinggo sp.flushFromLastCheckpoint(ctx)
, which may lead to thecancel
being executed immediately, not allowing the expected persister.gracePeriod(30-second timeout) to function correctly.https://github.com/pipe-cd/pipecd/blob/master/pkg/app/piped/logpersister/persister.go#L159 Why is this needed: