Insert query slowness
From: yudhi s <learnerdatabase99_at_gmail.com>
Date: Sat, 29 Oct 2022 04:04:19 +0530
Message-ID: <CAEzWdqeTXJLUhrfOOpFVKbPrgEGnmFfbwszNz50QNVDbQna65g_at_mail.gmail.com>
Hello Experts, It's version 19.15 of the oracle. But the OFE has still been set as 11.2.0.4. We are seeing after our infra team applied april 2022 database and grid patch(below is the exact detail), almost all of the conventional batch insert queries are running slower(e.g. ~5ms/execution before patch vs 10ms/execution now). The main time component for this query is shown to be "CPU time" in dba_hist_sqlstat, which has been increased. However, when checking the wait event pattern from dba_hist_system_event, the response time , we see for 'cell single block physical read' wait event the response time , has been increased up from <1ms before patch to ~10ms+ now after patch. Also we are seeing a new wait event "cell single block physical read: flash cache" which was not there before the patch. Want to know if anyone has encountered such an issue? And this symptom we are seeing across multiple databases in the same cluster.There is nothing much in the insert queries. The insert queries, looks like "Insert into tab1(c1,c2,....).. values(c1,c2,....); "
Date: Sat, 29 Oct 2022 04:04:19 +0530
Message-ID: <CAEzWdqeTXJLUhrfOOpFVKbPrgEGnmFfbwszNz50QNVDbQna65g_at_mail.gmail.com>
Hello Experts, It's version 19.15 of the oracle. But the OFE has still been set as 11.2.0.4. We are seeing after our infra team applied april 2022 database and grid patch(below is the exact detail), almost all of the conventional batch insert queries are running slower(e.g. ~5ms/execution before patch vs 10ms/execution now). The main time component for this query is shown to be "CPU time" in dba_hist_sqlstat, which has been increased. However, when checking the wait event pattern from dba_hist_system_event, the response time , we see for 'cell single block physical read' wait event the response time , has been increased up from <1ms before patch to ~10ms+ now after patch. Also we are seeing a new wait event "cell single block physical read: flash cache" which was not there before the patch. Want to know if anyone has encountered such an issue? And this symptom we are seeing across multiple databases in the same cluster.There is nothing much in the insert queries. The insert queries, looks like "Insert into tab1(c1,c2,....).. values(c1,c2,....); "
33815596;OCW RELEASE UPDATE 19.15.0.0.0 (33815596) 33806152;Database Release Update : 19.15.0.0.220419 (33806152) 33803476 - Grid patch
Regards
Yudhi
-- http://www.freelists.org/webpage/oracle-lReceived on Sat Oct 29 2022 - 00:34:19 CEST