xlog patches reviewed

"Mikheev, Vadim" <vmikheev@SECTORBASE.COM> writes:

but it's hard to notice eg misprints in 44K file -:)
I think we should apply patches and hard test
recovering for a few days (power off/pg_ctl -m i stop
with dozens update transactions).

OK. I haven't finished putting together an xlog-reset utility quite
yet, but I will go ahead and apply what I have.

CAUTION TO ONLOOKERS: if you update from CVS after I make this patch,
you will need to initdb!! Wait around for the log-reset utility if you
are running a database you don't want to initdb. I should have that in
another day or so.

regards, tom lane

On Mon, 12 Mar 2001, Mikheev, Vadim wrote:

but it's hard to notice eg misprints in 44K file -:)
I think we should apply patches and hard test
recovering for a few days (power off/pg_ctl -m i stop
with dozens update transactions).

if this is the case, can we look at applying that patch tonight, give ppl
till Friday to test and put out a RC1 depending on the results?

but it's hard to notice eg misprints in 44K file -:)
I think we should apply patches and hard test
recovering for a few days (power off/pg_ctl -m i stop
with dozens update transactions).

if this is the case, can we look at applying that patch
tonight, give ppl till Friday to test and put out a RC1
depending on the results?

I think so.

Vadim

if this is the case, can we look at applying that patch
tonight, give ppl till Friday to test and put out a RC1
depending on the results?

Patch committed. There are still some loose ends to clean up:

* I need to finish making an xlog-reset utility for contrib.

* I stubbed out shmctl(IPC_STAT) in the BeOS and QNX4 emulations of
SysV shared memory (src/backend/port/beos/shm.c,
src/backend/port/qnx4/shm.c). This means that the new code to detect
postmaster-dead-but-old-backends-still-running will never detect any
problem on those platforms. Perhaps people who use those platforms
can test and contribute real implementations?

However, these shouldn't affect testing.

regards, tom lane

On Mon, 12 Mar 2001, Tom Lane wrote:

if this is the case, can we look at applying that patch
tonight, give ppl till Friday to test and put out a RC1
depending on the results?

Patch committed. There are still some loose ends to clean up:

* I need to finish making an xlog-reset utility for contrib.

* I stubbed out shmctl(IPC_STAT) in the BeOS and QNX4 emulations of
SysV shared memory (src/backend/port/beos/shm.c,
src/backend/port/qnx4/shm.c). This means that the new code to detect
postmaster-dead-but-old-backends-still-running will never detect any
problem on those platforms. Perhaps people who use those platforms
can test and contribute real implementations?

However, these shouldn't affect testing.

Great, then let's go with a RC1 on Friday and see if we can get 7.1 out in
'02 :)

The Hermit Hacker writes:

if this is the case, can we look at applying that patch tonight, give ppl
till Friday to test and put out a RC1 depending on the results?

This should probably be called beta6, given that there is a lot of new
code in it.

--
Peter Eisentraut peter_e@gmx.net http://yi.org/peter-e/