I faced a similar situation sometime back in postgres 7.3.4 and i did
the following steps and got the problem fixed.
1. Increase max_fsm_pages (may need to increase max_fsm_relations also
if you have many objects)
2. Reindex the indexes which have grown very big (find them using
pg_class table)
3
Patrice Beliveau <[EMAIL PROTECTED]> writes:
> I got
> ERROR: cache lookup failed for type 12345
Show us an *exact* procedure to reproduce this, and mention exactly
which PG version you saw it on.
regards, tom lane
---(end of broadcast)
I'm doing a very simple thing...creating a schema xyz;creating domain xyz.abc; (about 100)creating table xyz.def' (abour 100)thendrop schema xyz cascade;after a list ofNOTICE: drop cascade to type xyz.aaa...I gotERROR: cache lookup failed for type 12345I thing that maybe there is(are) parameter(s)
did you ever get past your problem? I saw a post in a search, but
never really saw the solution. original link I found:
http://archives.postgresql.org/pgsql-admin/2005-07/msg00166.php
I (believe) the correct solution for postgres on gentoo linux is:
mkdir /var/lib/postgresql
chown po
Help me!
I have edited the program testlibpq.c which is copied from the
PostgreSQL document to make it become a CGI program. And I have
succeeded to compile the programme, testlibpq.c from the command line:
[EMAIL PROTECTED] src]$ cc -o testlibpq testlibpq.c -lpq
And it can run from the com
I loaded a 7.4.x dump into a new 8.1.1 database and found out what happens
if you had the same name as both a user and a group. You can get users with
more rights than they had before. I guess it is too late, but perhaps a
mention in the release text would have been a good idea. Advise people to
r
Donald Fraser wrote:
> We developed and have been using since 7.1, and currently running 7.4,
> bespoke client / database software. The ability to manage users and security
> was of high priority and we therefore developed a much more elaborate user
> definition where by the information about user
"Donald Fraser" <[EMAIL PROTECTED]> writes:
> Yes we do have another use.
> We developed and have been using since 7.1, and currently running 7.4,
> bespoke client / database software. The ability to manage users and security
> was of high priority and we therefore developed a much more elaborate u
Jonathan Parkin <[EMAIL PROTECTED]> writes:
> I have a reasonably large, live, system-critical database. A perl
> script on another machine connects and issues a sequence of commands in
> a transaction, the last of which is a delete. The delete never returns
> a response, and the connection never
- Original Message -
From: "Tom Lane"
> "Donald Fraser" writes:
> > 1) Was there a good reason for dropping this feature?
> Users/groups are identified by OIDs now, not by sysids.
>
> The original motivation for allowing manual assignment of userids was to
> allow recovery if a user w
I have a reasonably large, live, system-critical database. A perl
script on another machine connects and issues a sequence of commands in
a transaction, the last of which is a delete. The delete never returns
a response, and the connection never times out. The postgres process
handling the delet
"Donald Fraser" <[EMAIL PROTECTED]> writes:
> 1) Was there a good reason for dropping this feature?
Users/groups are identified by OIDs now, not by sysids.
The original motivation for allowing manual assignment of userids was to
allow recovery if a user was dropped while still owning objects ---
In the latest version of PostgreSQL (8.1.x) the
ability to set the system id (oid) (option: WITH SYSID) when creating a user has
been dropped.
I have two questions:
1) Was there a good reason for dropping this
feature? (people complain when you take things away from them)
2) What problems if
Chris Hoover <[EMAIL PROTECTED]> writes:
> We are running a typical OLTP system with some reporting done against it.
> What value would you recommend as a starting point for the work_mem.
You probably want small work_mem for the transactional sessions (4Mb
doesn't sound out of line, maybe even le
On 12/19/05, Tom Lane <[EMAIL PROTECTED]> wrote:
> Chris Hoover <[EMAIL PROTECTED]> writes:
> > We are in the process of upgrading our 7.3.4 servers to 8.1, and I
> > wanted to have a sanity check ran against my config file. Do you see
> > any issues with these configs?
>
> > work_mem = 4096
Boguk Maxim <[EMAIL PROTECTED]> writes:
> I trying understand why vacuum analyze do not work at all.
I think either you're not actually applying vacuum analyze at all
(better double-check your daily maintenance script), or you're somehow
managing to apply it at a time when it can't remove old rows
But why i should 'vacuum full' each day? All what i read suggest use vacuum
analyze once per day for such tables.
Again: i doing vacuum analyze each time after update table (once per 24
hour) (all other time table is read only really)
(and i see no reason for vacuum the table more frequently if i
The table events_extra needs to be full vacuumed to free the disk space with respect to this table and vacuum the table more frequently .
You need to reindex the table once in a week ... (it depends on growth of the index size )
12/20/05, Boguk Maxim <[EMAIL PROTECTED]> wrote:
I have o
> we made some custom types (written in C) for PostgreSQL onto
> a Linux server.
> We need to deploy a Windows server...
>
> The documentation does not talk about a shared library for Windows
> (http://www.postgresql.org/docs/8.1/interactive/xfunc-c.html#DFUNC)
> I tried to compile its custom ty
Hi there,
we made some custom types (written in C) for PostgreSQL onto a Linux
server.
We need to deploy a Windows server...
The documentation does not talk about a shared library for Windows
(http://www.postgresql.org/docs/8.1/interactive/xfunc-c.html#DFUNC)
I tried to compile its custom t
Le mardi 20 décembre 2005 à 13:20 +0300, Boguk Maxim a écrit :
> Try add that into postgresql.conf:
>
> add_missing_from = on
Thanks that did the trick but it does not explain why the same query
works on the other page... =;-D
Tony
--
Tony Grant
www.tgds.net - vente d'ordinateurs mini-itx en li
Try add that into postgresql.conf:
add_missing_from = on
More details look here:
http://www.postgresql.org/docs/8.1/interactive/runtime-config-compatible.htm
l
(
add_missing_from (boolean)
When on, tables that are referenced by a query will be automatically added
to the FROM clause if not already
Ok this is probably a JDBC thing but as many admins seem to be running
webapps and it is a version migration problem I shall ask here:
I have migrated the backend from 7.2.1 to 8.1.1
Most of the JSPs work just fine several others don't
§
org.postgresql.util.PSQLException: ERREUR:
I have one small table (like 35000 records) in my database.
Table updated once per day (1%-5% records updated 0.1%-2% new records
inserted).
(parent table updating at same time... no other updates/deletes on tables
events and events_extra ever doing).
For such usage mode i choose 'vacuum analyze' t
Jim C. Nasby wrote:
I believe that by default the custom format is gzip'd, which also means
you probably couldn't exclude just the timestamp. Though if you're using
rsync, I think there's a good chance it would only tranfer the first few
hundred KB...
Yes that is true.
What I forgot to mention
25 matches
Mail list logo
