Tim Armstrong has posted comments on this change. ( http://gerrit.cloudera.org:8080/14332 )
Change subject: IMPALA-8962: FETCH_ROWS_TIMEOUT_MS should apply before rows are available ...................................................................... Patch Set 3: (2 comments) Nice tests! http://gerrit.cloudera.org:8080/#/c/14332/3/be/src/runtime/coordinator.cc File be/src/runtime/coordinator.cc: http://gerrit.cloudera.org:8080/#/c/14332/3/be/src/runtime/coordinator.cc@687 PS3, Line 687: parent_request_state_->fetch_rows_timeout_us() : - parent_request_state_->block_on_wait_time_us If by coincidence block_on_wait_time_us_ is the same as fetch_rows_timeout_us, can't this work out to 0 and incorrectly block indefinitely? http://gerrit.cloudera.org:8080/#/c/14332/3/be/src/service/client-request-state.h File be/src/service/client-request-state.h: http://gerrit.cloudera.org:8080/#/c/14332/3/be/src/service/client-request-state.h@353 PS3, Line 353: uint64_t block_on_wait_time_us_ = 0; I think storing this per-query is incorrect since there could be multiple concurrent calls to BlockOnWait() and one thread might pick up the other thread's value. I think that might be an unimportant edge case and we could probably live with that but I think it still makes the logic unnecessarily confusing because this value really fits with the RPC request, not the query. I could see either threading it through the function calls (probably the most obvious and explicit way) or using a thread-local variable. -- To view, visit http://gerrit.cloudera.org:8080/14332 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: I2cba6bf062dcc1af19471d21857caa797c1ea4a4 Gerrit-Change-Number: 14332 Gerrit-PatchSet: 3 Gerrit-Owner: Sahil Takiar <[email protected]> Gerrit-Reviewer: Impala Public Jenkins <[email protected]> Gerrit-Reviewer: Sahil Takiar <[email protected]> Gerrit-Reviewer: Thomas Tauber-Marshall <[email protected]> Gerrit-Reviewer: Tim Armstrong <[email protected]> Gerrit-Comment-Date: Wed, 02 Oct 2019 20:13:56 +0000 Gerrit-HasComments: Yes
