Jesse Norell wrote:
Matthew T. O'Connor wrote:
I assume the "fix" you are talking about is that starting with
PostgreSQL 7.4 Vacuum reclaims index space. That resolves the major
remaining file bloat problem with PostgreSQL. As of 7.4 as long as your
FSM setting is high enough, and you v
Hi,
please, correct me if I am on the wrong list. I think, that everything
related to current cvs version belongs to dbmail.devel?
today I checked out current cvs version. It take some time for me to
convert tables. It works now, but fetching every single mail is very slow.
Here is a quota from
Sergey Spiridonov wrote:
problem is in query: SELECT messageblk FROM dbmail_messageblks LEFT JOIN
dbmail_messages USING (physmessage_id) WHERE
dbmail_messages.message_idnr = '551047' ORDER BY messageblk_idnr
which takes up to 6 seconds!!!
Slightly rewritten query
SELECT messageblk FROM dbmail
Sergey Spiridonov <[EMAIL PROTECTED]> said:
> Sergey Spiridonov wrote:
>> problem is in query: SELECT messageblk FROM dbmail_messageblks LEFT JOIN
>> dbmail_messages USING (physmessage_id) WHERE
>> dbmail_messages.message_idnr = '551047' ORDER BY messageblk_idnr
>>
>> which takes up to 6 seconds!
Does this mean postgres has trouble with left joins. Afaik, both queries are
equivalent. Mysql treats them exactly the same, according to the docs and analyze.
Or could this be related to the index bloat Matthew and Jesse have been
discussing?
Sergey Spiridonov wrote:
Sergey Spiridonov wrote
Aaron Stone wrote:
I get the same results for both queries, however I also get the same
EXPLAIN. I'm using MyISAM tables with the earlier 2.0 defs renamed to
dbmail_ by hand. Which database scripts did you use? I can run them on a
test database on my machine and see if there are any differences
Sergey Spiridonov wrote:
Seems to be postgresql problem. This query is used in 3 places:
2 times in dbmsgbuf.c and 1 time in dbsearch.c. I changed them and
till now everything works fine...
# postgres --version
postgres (PostgreSQL) 7.3.6
This query is a good candidate for being replaced th
> This is very interesting. Would you mind writing a README.postgresql
> with this (and possibly more) information for tuning
> PostgreSQL for DbMail?
I wouldn't mind at all. However if there is someone that feels they may
be more qualified, then by all means step in. I'm still pretty new to
dbm
Sergey Spiridonov wrote:
> I converted to dbmail_* by hand just yesterday also. I'm using
> postgres, and for postgres explain differs:
This is for Postgres 7.4.2:
> dbmail=# explain SELECT messageblk FROM dbmail_messageblks LEFT JOIN
> dbmail_messages USING (physmessage_id) WHERE
> dbmail_mes
Could someone first please confirm that left joins are really this broken in
postgres. Perhaps it's something in Sergey's setup/build of postgres.
Left joins are *so* much more readable than normal joins.
Ilja Booij wrote:
Sergey Spiridonov wrote:
Seems to be postgresql problem. This query
Thomas Mueller wrote:
Sergey Spiridonov wrote:
I converted to dbmail_* by hand just yesterday also. I'm using
postgres, and for postgres explain differs:
This is for Postgres 7.4.2:
O.K. it is time to upgrade for me. Thank you.
--
Best regards, Sergey Spiridonov
[EMAIL PROTECTED] wrote:
> Could someone first please confirm that left joins are really
> this broken in
> postgres. Perhaps it's something in Sergey's setup/build of postgres.
Funny, I get the same result for both queries. Postgres 7.4.3 on
FreeBSD...
creating dbmail-pop3d
gcc -DHAVE_CONFIG_H -I. -I. -I.-fomit-frame-pointer -g -O2 -W -Wall
-Wpointer-arith -Wstrict-prototypes -c quota.c
gcc -DHAVE_CONFIG_H -I. -I. -I.-fomit-frame-pointer -g -O2 -W -Wall
-Wpointer-arith -Wstrict-prototypes -c imap4.c
gcc -DHAVE_CONFIG_H -I. -I. -I.-f
Oops. My wrong.
I could've sworn I did a compile before committing.. but it seems I didn't..
it's fixed now.
Ilja
Igor Olemskoi wrote:
creating dbmail-pop3d
gcc -DHAVE_CONFIG_H -I. -I. -I.-fomit-frame-pointer -g -O2 -W -Wall
-Wpointer-arith -Wstrict-prototypes -c quota.c
gcc -DHAVE_CONF
Donald C. Sumbry ][ wrote:
[EMAIL PROTECTED] wrote:
Could someone first please confirm that left joins are really
this broken in
postgres. Perhaps it's something in Sergey's setup/build of postgres.
Funny, I get the same result for both queries. Postgres 7.4.3 on
FreeBSD...
Sergey Spiridonov <[EMAIL PROTECTED]> writes:
> Yes, but for me it makes no sense. I would understand, if it will be
>
> SELECT messageblk FROM dbmail_messages LEFT JOIN dbmail_messageblks
>USING (physmessage_id)
>WHERE dbmail_messages.message_idnr = '%llu'
>
> s
Sergey Spiridonov wrote:
Sergey Spiridonov <[EMAIL PROTECTED]> writes:
Yes, but for me it makes no sense. I would understand, if it will be
SELECT messageblk FROM dbmail_messages LEFT JOIN dbmail_messageblks
USING (physmessage_id)
WHERE dbmail_messages.message_i
> > We do regularly drop an rebuild indexes on messages, but it looks like
> >we do not on messageblks. There is one index on messageblks(message_idnr)
> >and the primary key, both of which are bigint's (8 bytes). So with
> >max(messageblk_idnr) at < 45million, that should account for about 72
Hello,
mysql> SELECT * FROM table1,table2 WHERE table1.id=table2.id;
mysql> SELECT * FROM table1 LEFT JOIN table2 ON table1.id=table2.id;
mysql> SELECT * FROM table1 LEFT JOIN table2 USING (id);
These are all semantically identical in mysql.
No, these are not all semantically e
19 matches
Mail list logo
