| From: | Gunther Schadow <gunther(at)aurora(dot)regenstrief(dot)org> |
|---|---|
| To: | Ian Cass <ian(dot)cass(at)mBlox(dot)com> |
| Cc: | pgsql-sql(at)postgresql(dot)org |
| Subject: | Re: select failure |
| Date: | 2002-06-09 20:01:42 |
| Message-ID: | 3D03B426.6070900@aurora.regenstrief.org |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-sql |
Ian Cass wrote:
> Debian apt-get installed 7.2.1. No I don't have ppl deleting files.
>
> What do I need to do to fix it?
>
> Worst case, I can probably remove the table and rebuild it. Although it
> contains about 27million rows and will take 2 or 3 days to rebuild.
>
> Thing is, how did it happen and is it likely to happen again?
I don't know. But something is really in very bad shape. I suggest
you backtrack one minor version to 7.2.
Rebuilding the entire thing is what I would do. However, I don't
really undestand what that pg_clog subdirectory is. In 2.7 that
I use I recall that all data is in the data/base directory just
as it was since time began. If I ever had a file vanish from there
I would definitely completely rebuild the database (from backups
if any or from scratch.)
Rebuilding 27million rows shouldn't take 2 or 3 days. Are you
sure you do a COPY ... FROM instead of individual INSERTs? Also
you may consider turning off indexes and deferring foreign key
triggers. However, I don't need to do this to load a similar
sized database overnight. YMMV, depends on machine speed too,
of course.
Good luck, and beware, I'm not really the pgsql geek who really
understands your error. I just use common sense, so, don't take
my word for gold w/r/t the recommendation to backtrack version and
rebuild.
regards,
-Gunther
>
> --
> Ian Cass
>
> ----- Original Message -----
> From: "Gunther Schadow" <gunther(at)aurora(dot)regenstrief(dot)org>
> To: "Ian Cass" <ian(dot)cass(at)mblox(dot)com>
> Cc: <pgsql-sql(at)postgresql(dot)org>
> Sent: Sunday, June 09, 2002 7:10 PM
> Subject: Re: [SQL] select failure
>
>
>
>>Ian Cass wrote:
>>
>>
>>>Ideas where I should look for the answer to this problem & how I should
>>>
> fix
>
>>>it?
>>>
>>>observer=# select host, current_route, count(current_route)
>>>observer-# from messages_200205
>>>observer-# group by host, current_route;
>>>FATAL 2: open of /var/lib/postgres/data/pg_clog/0534 failed: No such
>>>
> file
>
>>>or directory
>>>
>>
>>OUCH! that looks quite bad. It's a postgresql internal error.
>>What version of PostgreSQL do you run? Is this some beta version?
>>Or do you have someone deleting files randomly?
>>
>>regards,
>>-Gunther
>>
>>
>>
>>--
>>Gunther Schadow, M.D., Ph.D. gschadow(at)regenstrief(dot)org
>>Medical Information Scientist Regenstrief Institute for Health Care
>>Adjunct Assistant Professor Indiana University School of Medicine
>>tel:1(317)630-7960 http://aurora.regenstrief.org
>>
>>
>>
>>---------------------------(end of broadcast)---------------------------
>>TIP 2: you can get off all lists at once with the unregister command
>> (send "unregister YourEmailAddressHere" to majordomo(at)postgresql(dot)org)
>>
>>
--
Gunther Schadow, M.D., Ph.D. gschadow(at)regenstrief(dot)org
Medical Information Scientist Regenstrief Institute for Health Care
Adjunct Assistant Professor Indiana University School of Medicine
tel:1(317)630-7960 http://aurora.regenstrief.org
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Ian Cass | 2002-06-09 20:13:46 | Re: select failure |
| Previous Message | Tom Lane | 2002-06-09 19:28:07 | Re: select failure |