Details
-
Bug
-
Status: Open (View Workflow)
-
TBD
-
Resolution: Unresolved
-
31.1.0
-
None
-
CP: Roadmap backlog
-
Core: Platform
-
GBV
Description
Implement waitQueueTimeout for vertx-sql-client ConnectionPool.
The connection pool has a max size. RMB has DB_MAXPOOLSIZE to configure it.
If there are more parallel requests that require a connection they are put into the
wait queue. If connections in use take long or are not released after use the requests stay in the wait queue forever. The clients of the waiting requests do not get any feedback. They are stuck. The module is unresponsive.
Solution: Add a waitQueueTimeout to ConnectionPool. A timeout failure is returned if the request stays in the pool longer than waitQueueTimeout. Add DB_WAITQUEUETIMEOUT to RMB.
Workaround: Limit the number requests waiting in the wait queue: https://vertx.io/docs/apidocs/io/vertx/sqlclient/PoolOptions.html#setMaxWaitQueueSize-int-
TestRail: Results
Attachments
Issue Links
- is duplicated by
-
RMB-710 Timeout when exceeding DB_MAXPOOLSIZE
-
- Closed
-