RE: Wrong statistics for size of XLOG_SWITCH during pg_waldump.

From: <Shinya11(dot)Kato(at)nttdata(dot)com>
To: <movead(dot)li(at)highgo(dot)ca>
Cc: <pgsql-hackers(at)postgresql(dot)org>, <andres(at)anarazel(dot)de>, <michael(at)paquier(dot)xyz>, <ahsan(dot)hadi(at)highgo(dot)ca>, <horikyota(dot)ntt(at)gmail(dot)com>
Subject: RE: Wrong statistics for size of XLOG_SWITCH during pg_waldump.
Date: 2020-12-04 04:20:47
Message-ID: 240cd11dc6664dfcbef50c2d6b16ccb2@MP-MSGSS-MBX001.msg.nttdata.co.jp
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

When I execute pg_waldump, I found that XLOG/SWITCH_JUNK appears twice.
Is this problem solved by the way of correcting the previously discussed Transaction/COMMIT?

$ ../bin/pg_waldump --stats=record ../data/pg_wal/000000010000000000000001
Type N (%) Record size (%) FPI size (%) Combined size (%)
---- - --- ----------- --- -------- --- ------------- ---
XLOG/CHECKPOINT_SHUTDOWN 5 ( 0.01) 570 ( 0.01) 0 ( 0.00) 570 ( 0.01)
XLOG/CHECKPOINT_ONLINE 6 ( 0.02) 684 ( 0.02) 0 ( 0.00) 684 ( 0.01)
XLOG/NEXTOID 3 ( 0.01) 90 ( 0.00) 0 ( 0.00) 90 ( 0.00)
XLOG/FPI 290 ( 0.80) 14210 ( 0.34) 638216 ( 40.72) 652426 ( 11.30)
Transaction/COMMIT 12 ( 0.03) 408 ( 0.01) 0 ( 0.00) 408 ( 0.01)
Transaction/COMMIT 496 ( 1.36) 134497 ( 3.20) 0 ( 0.00) 134497 ( 2.33)
Storage/CREATE 13 ( 0.04) 546 ( 0.01) 0 ( 0.00) 546 ( 0.01)
CLOG/ZEROPAGE 1 ( 0.00) 30 ( 0.00) 0 ( 0.00) 30 ( 0.00)
Database/CREATE 2 ( 0.01) 84 ( 0.00) 0 ( 0.00) 84 ( 0.00)
Standby/LOCK 142 ( 0.39) 5964 ( 0.14) 0 ( 0.00) 5964 ( 0.10)
Standby/RUNNING_XACTS 13 ( 0.04) 666 ( 0.02) 0 ( 0.00) 666 ( 0.01)
Standby/INVALIDATIONS 136 ( 0.37) 12416 ( 0.30) 0 ( 0.00) 12416 ( 0.22)
Heap2/CLEAN 132 ( 0.36) 8994 ( 0.21) 0 ( 0.00) 8994 ( 0.16)
Heap2/FREEZE_PAGE 245 ( 0.67) 168704 ( 4.01) 0 ( 0.00) 168704 ( 2.92)
Heap2/CLEANUP_INFO 2 ( 0.01) 84 ( 0.00) 0 ( 0.00) 84 ( 0.00)
Heap2/VISIBLE 424 ( 1.16) 25231 ( 0.60) 352256 ( 22.48) 377487 ( 6.54)
XLOG/SWITCH_JUNK 0 ( 0.00) 0 ( 0.00) 0 ( 0.00) 0 ( 0.00)
Heap2/MULTI_INSERT 1511 ( 4.15) 287727 ( 6.84) 12872 ( 0.82) 300599 ( 5.21)
Heap2/MULTI_INSERT+INIT 46 ( 0.13) 71910 ( 1.71) 0 ( 0.00) 71910 ( 1.25)
Heap/INSERT 8849 ( 24.31) 1288414 ( 30.62) 25648 ( 1.64) 1314062 ( 22.76)
Heap/DELETE 25 ( 0.07) 1350 ( 0.03) 0 ( 0.00) 1350 ( 0.02)
Heap/UPDATE 173 ( 0.48) 55238 ( 1.31) 5964 ( 0.38) 61202 ( 1.06)
Heap/HOT_UPDATE 257 ( 0.71) 27585 ( 0.66) 1300 ( 0.08) 28885 ( 0.50)
XLOG/SWITCH_JUNK 0 ( 0.00) 0 ( 0.00) 0 ( 0.00) 0 ( 0.00)
Heap/LOCK 180 ( 0.49) 9800 ( 0.23) 129812 ( 8.28) 139612 ( 2.42)
Heap/INPLACE 214 ( 0.59) 44520 ( 1.06) 40792 ( 2.60) 85312 ( 1.48)
Heap/INSERT+INIT 171 ( 0.47) 171318 ( 4.07) 0 ( 0.00) 171318 ( 2.97)

Regards,
Shinya Kato

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Amit Kapila 2020-12-04 04:59:42 Re: Single transaction in the tablesync worker?
Previous Message Tang, Haiying 2020-12-04 03:42:19 RE: [Patch] Optimize dropping of relation buffers using dlist