Hi,
I noticed an issue in the pgbench progress message where an extra
closing parenthesis )) appears, as shown below:
700 of 1000 tuples (70%) of pgbench_accounts done (elapsed 19.75
s, remaining 8.46 s))
This occurs when running commands like pgbench -i -s100 and is caused by
left
the return values of consume_xids and
consume_xids_until, and the rationale for incrementing consumed++ when
GetTopFullTransactionId() is called.
Also, I made another patch (support_blksz_32k.diff) that supports the
block size == 32K case.
Best,
Yushi Ogiwara
diff --git a/src/test/modules
k this is not the case since XidSkip returns min(UINT32_MAX - 5 -
low, *), which prevents the wrap-around of nextXid.
Regards,
Yushi Ogiwara
, Oct 14, 2024 at 6:51 PM Yushi Ogiwara
wrote:
Hi,
I found that the consume_xids function incorrectly advances XIDs as
shown:
postgres=# select txid_current();
txid_current
--
746
(1 row)
postgres=# select consume_xids('100');
consume_xids
--
tgres=*# select consume_xids('100');
consume_xids
--
1846
(1 row)
Regards,
YushiFrom 59bc3f300b8447a407c8f90ebb542d1c0ef0e932 Mon Sep 17 00:00:00 2001
From: Yushi Ogiwara
Date: Fri, 11 Oct 2024 14:45:26 +0900
Subject: [PATCH] fix xid
---
src/test/modules/xid_wraparou
601c3510cb77c99abde03722ef72ba77862c40f2 Mon Sep 17 00:00:00 2001
From: Yushi Ogiwara
Date: Fri, 11 Oct 2024 15:07:08 +0900
Subject: [PATCH] fix attr
---
src/test/modules/xid_wraparound/xid_wraparound--1.0.sql | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/src/test/modules