Tom Lane wrote:
Some time ago, Jon Lapham <[EMAIL PROTECTED]> wrote:
Today I had a power outage which upon reboot seems to have done
something to cause Postgresql to not restart properly. This has
happened to me before:
http://archives.postgresql.org/pgsql-general/2006-09/msg00938.php
We fi
Some time ago, Jon Lapham <[EMAIL PROTECTED]> wrote:
> Today I had a power outage which upon reboot seems to have done
> something to cause Postgresql to not restart properly. This has
> happened to me before:
> http://archives.postgresql.org/pgsql-general/2006-09/msg00938.php
We finally tracke
Tom Lane wrote:
BTW, you said you were on FC6, but what kernel version exactly?
(uname -a is good)
[EMAIL PROTECTED] > uname -a
Linux bilbo.localdomain 2.6.19-1.2911.6.5.fc6 #1 SMP Sun Mar 4 16:05:34
EST 2007 x86_64 x86_64 x86_64 GNU/Linux
[EMAIL PROTECTED] > cat /proc/cpuinfo
processor
BTW, you said you were on FC6, but what kernel version exactly?
(uname -a is good)
regards, tom lane
---(end of broadcast)---
TIP 9: In versions below 8.0, the planner will ignore your desire to
choose an index scan if
Jon Lapham <[EMAIL PROTECTED]> writes:
> Tom Lane wrote:
>> Just for completeness, could we see the output of "ipcs -a" now?
> [EMAIL PROTECTED] ~]# ipcs -a
Thanks. The postgres segment looks pretty standard here. I'm a bit
curious what all the lapham-owned destroyed segments are --- any idea?
A
Tom Lane wrote:
Jon Lapham <[EMAIL PROTECTED]> writes:
Tom Lane wrote:
It would seem that maybe your kernel has a different idea of what EIDRM
means than we do. You say this persists across a reboot?
I probably should have asked before doing this... but I just rebooted
and postgresql came u
Jon Lapham <[EMAIL PROTECTED]> writes:
> Tom Lane wrote:
>> It would seem that maybe your kernel has a different idea of what EIDRM
>> means than we do. You say this persists across a reboot?
> I probably should have asked before doing this... but I just rebooted
> and postgresql came up fine.
Tom Lane wrote:
Oh, that's interesting. The code is barfing on this because
/*
* Otherwise, we had better assume that the segment is in use. The
* only likely case is EIDRM, which implies that the segment has been
* IPC_RMID'd but there are still processes att
Jon Lapham <[EMAIL PROTECTED]> writes:
> chdir("/var/lib/pgsql/data")= 0
> open("postmaster.pid", O_RDWR|O_CREAT|O_EXCL, 0600) = -1 EEXIST (File
> exists)
> open("postmaster.pid", O_RDONLY)= 3
> read(3, "2809\n/var/lib/pgsql/data\n 54320"..., 1123) = 45
> close(3)
Tom Lane wrote:
Jon Lapham <[EMAIL PROTECTED]> writes:
Tom Lane wrote:
Let's see the output from
sudo ipcs -a
[EMAIL PROTECTED] ~]# ipcs -a
[ no postgres-owned segments... ]
Well, that's dang interesting. Could you run the postmaster under
strace and send the output?
Let me know if thi
Jon Lapham <[EMAIL PROTECTED]> writes:
> Tom Lane wrote:
>> Let's see the output from
>> sudo ipcs -a
> [EMAIL PROTECTED] ~]# ipcs -a
[ no postgres-owned segments... ]
Well, that's dang interesting. Could you run the postmaster under
strace and send the output?
regards,
Tom Lane wrote:
Jon Lapham <[EMAIL PROTECTED]> writes:
What can I do to help diagnose what is happening?
Let's see the output from
ps auxww | grep postgres
sudo ipcs -a
[EMAIL PROTECTED] ~]# ps auxww | grep postgres
root 3969 0.0 0.0 60272 688 pts/1R+ 15:47 0:00 grep
postg
Jon Lapham <[EMAIL PROTECTED]> writes:
> What can I do to help diagnose what is happening?
Let's see the output from
ps auxww | grep postgres
sudo ipcs -a
(or local equivalents)
regards, tom lane
---(end of broadcast)---
TI
Today I had a power outage which upon reboot seems to have done
something to cause Postgresql to not restart properly. This has
happened to me before:
http://archives.postgresql.org/pgsql-general/2006-09/msg00938.php
...and during this previous discussion, Tom Lane pointed out that I may
have
Tom Lane wrote:
Jon Lapham <[EMAIL PROTECTED]> writes:
[...]
-- Shared Memory Segments
keyshmid owner perms bytes nattch status
0x 0 root 77794208 0
0x6a6b6cbd 53411843 lapham6003840
0x12ac1925
Jon Lapham <[EMAIL PROTECTED]> writes:
> Is it important for you to know that at the time of the power outage, I
> *did* have 2 closed source kernel modules loaded, vmware's and NVidia's.
> (This is a development machine, not production...). Could one of
> these modules screwed up somehow and
Tom Lane wrote:
Jon Lapham <[EMAIL PROTECTED]> writes:
To stress test, I turned off the power while actively processing db
operations, such as: loading data into a database, create a new
database, delete contents of a large table within a transaction.
Anyway, in every case I could not reproduce
Harald Armin Massa wrote:
It is inherently flawed. VMware really "powers down", that is, the
operating system has time to shut down. Or, in other incarnations,
VMware freezes the system state.
It's nothing near a real power outage, which gives no time for anything.
On my VMware window, there
Tom Lane wrote:
Jon Lapham <[EMAIL PROTECTED]> writes:
Is it important for you to know that at the time of the power outage, I
*did* have 2 closed source kernel modules loaded, vmware's and NVidia's.
(This is a development machine, not production...). Could one of
these modules screwed up s
Jon Lapham <[EMAIL PROTECTED]> writes:
> To stress test, I turned off the power while actively processing db
> operations, such as: loading data into a database, create a new
> database, delete contents of a large table within a transaction.
> Anyway, in every case I could not reproduce the issue
Jon,For what it is worth, I created a FC5 VMware installation and loaded mydatabase data into it. I simulated a bunch of power outages by telling
VMware to power off the vm. Is this a good simulation of a poweroutage, or is there something inherently flawed about using a VM to testthis?It is inhe
Tom Lane wrote:
Adrian Klaver <[EMAIL PROTECTED]> writes:
On Sunday 24 September 2006 09:17 am, Tom Lane wrote:
Jon Lapham <[EMAIL PROTECTED]> writes:
FATAL: pre-existing shared memory block (key 5432001, ID 65536) is
still in use
This is extremely odd, because a shared memory block could no
I wrote:
> Also, if it is repeatable, "ipcs -m" output will be useful context.
I forgot to mention: on Linux it's important to run ipcs as root
(eg "sudo ipcs -m") else it will lie to you. An incomplete listing
is worse than useless.
regards, tom lane
---
Adrian Klaver <[EMAIL PROTECTED]> writes:
> On Monday 25 September 2006 02:48 pm, Tom Lane wrote:
>> I spent quite some time today trying to duplicate this failure (by
>> pulling the plug on an up-to-date Fedora Core 5 machine). No luck.
> Is there something I could do to help capture useful info
On Monday 25 September 2006 02:48 pm, Tom Lane wrote:
> Adrian Klaver <[EMAIL PROTECTED]> writes:
> > On Sunday 24 September 2006 09:17 am, Tom Lane wrote:
> >> Jon Lapham <[EMAIL PROTECTED]> writes:
> >>> FATAL: pre-existing shared memory block (key 5432001, ID 65536) is
> >>> still in use
> >>
>
Adrian Klaver <[EMAIL PROTECTED]> writes:
> On Sunday 24 September 2006 09:17 am, Tom Lane wrote:
>> Jon Lapham <[EMAIL PROTECTED]> writes:
>>> FATAL: pre-existing shared memory block (key 5432001, ID 65536) is
>>> still in use
>>
>> This is extremely odd, because a shared memory block could not
Jon Lapham wrote:
> [EMAIL PROTECTED] ~]# ps -A | grep -i post
> 30760 ?00:00:00 postmaster
> 30762 ?00:00:00 postmaster
> 30764 ?00:00:00 postmaster
> 30765 ?00:00:00 postmaster
> 30766 ?00:00:00 postmaster
>
> ...is that normal to see 5 of them running?
On Sunday 24 September 2006 09:17 am, Tom Lane wrote:
> Jon Lapham <[EMAIL PROTECTED]> writes:
> > I recently had another electrical power outage that left my machine
> > unable to restart postgresql. I had previously reported this a while
> > ago:
> >
> > http://archives.postgresql.org/pgsql-gene
Jon Lapham wrote:
[...]
I do not *think* I am running 2 postmasters.
[EMAIL PROTECTED] ~]# service postgresql stop
Stopping postgresql service: [ OK ]
[EMAIL PROTECTED] ~]# ps -A | grep -i post
[EMAIL PROTECTED] ~]# service postgresql start
Starting postgresql se
Tom Lane wrote:
Jon Lapham <[EMAIL PROTECTED]> writes:
I recently had another electrical power outage that left my machine
unable to restart postgresql. I had previously reported this a while ago:
http://archives.postgresql.org/pgsql-general/2005-04/msg01286.php
Anyway, because I have seen
This doesn't make sense to me. A reboot will absolutly kill any
existing shared memory blocks, how can it possibly be complaining about
it?
PostgreSQL complains if it finds a postmaster.pid. As far as I can tell
it doesn't have anything to do with shared memory except that we are
tracking in
Jon Lapham <[EMAIL PROTECTED]> writes:
> I recently had another electrical power outage that left my machine
> unable to restart postgresql. I had previously reported this a while ago:
> http://archives.postgresql.org/pgsql-general/2005-04/msg01286.php
> Anyway, because I have seen this problem
On Sun, Sep 24, 2006 at 10:11:00AM -0300, Jon Lapham wrote:
> I recently had another electrical power outage that left my machine
> unable to restart postgresql. I had previously reported this a while ago:
FWIW, I've crashed my machine a lot of times and never run into this
problem. However, I r
> ...but for
> yucks I decided to imagine I did not know. What would I need to do to
> figure this out.
> Maybe it is just a right
> of passage that users of postgresql will just have to learn about this.
I would imagine that I am one of the yucks like you. ;-) Anyway, everytime I
see a us
I recently had another electrical power outage that left my machine
unable to restart postgresql. I had previously reported this a while ago:
http://archives.postgresql.org/pgsql-general/2005-04/msg01286.php
Anyway, because I have seen this problem before, I knew exactly what the
solution to
35 matches
Mail list logo
