And with the timing escalations in SVN, we'll be able to see the queries
that take a very long time. Should I begin escalating sooner, like 5
seconds to INFO, 10 seconds to MESSAGE, 15 seconds to WARNING?
Aaron
On Thu, Jul 5, 2007, Jonathan Feally <[EMAIL PROTECTED]> said:
> Your problem is not with DBMAIL, but rather either a slow database
> server or really big mailbox. I have hacked my dbmail-imap to sleep for
> 120 seconds when doing a search and it caused this same error on SM
> 1.4.9a. The default timeout on the socket is 60 seconds and of course a
> query on a sql server on a huge mailbox such as looking for aa in all of
> the message bodies is going to take a long time, once it exceeds 60
> seconds, PHP fgets() will return because the socket says it timed out
> reading data. As suspected, it was a timeout issue, not a bug in DBMAIL.
>
> This patch below shows you where to add a line to change this timeout to
> help with these types of operations in SM, but if the mailbox is too
> big, the user will just complain that the email server stops working -
> copy/paste patch to file:
>
> --- functions/imap_general.php Wed Oct 4 12:35:42 2006
> +++ functions/imap_general.php Wed Jul 4 19:01:36 2007
> @@ -449,6 +449,9 @@
>
> $imap_stream = @fsockopen($imap_server_address, $imap_port,
> $error_number, $error_string, 15);
>
> + // Extend normal timeout from 60 seconds to 150 seconds for slow
> servers doing complex operations such as body searching
> + @stream_set_timeout($imap_stream, 150);
> +
> /* Do some error correction */
> if (!$imap_stream) {
> if (!$hide) {
>
>
> -Jon
>
>
> Jorge Bastos wrote:
>> PHP 5.2.1
>> SM 1.4.9
>> Apache 2.2.4
>>
>> The error it's from dbmail, doesn't seams to be an error from a timeout.
>>
>> ERROR:ERROR: Connection dropped by IMAP server.
>> Query: SEARCH CHARSET ISO-8859-1 ALL BODY {2} aa
>>
>>
>>
>> Paul,
>> Do you help me debuging this with gdb?
>>
>>
>>
>>
>> ----- Original Message ----- From: "Jonathan Feally"
>> <[EMAIL PROTECTED]>
>> To: "DBMail mailinglist" <[email protected]>
>> Sent: Wednesday, July 04, 2007 11:46 PM
>> Subject: Re: [Dbmail] Bug #579
>>
>>
>>> Well - all page requests in SM do a login, the work, then log out.
>>> Thus each transaction is a separate login session. Looking at the SM
>>> code for this error - It is displayed when the fgets on the
>>> $imap_stream fails. So aside from the version numbers I've already
>>> asked for, what version apache, PHP, and SM are you running? This
>>> error doesn't actually mean that the TCP connection was closed to the
>>> server, but rather that reading from the connection failed, perhaps
>>> due to a time out on reading from the stream.
>>>
>>> -Jon
>>>
>>> Jorge Bastos wrote:
>>>> Hum,
>>>> But that i don't know, but like i said in the previous email, after
>>>> i received the error, i pushed the logout button on squirrelmail.
>>>>
>>>>
>>>> ----- Original Message ----- From: "Paul J Stevens" <[EMAIL PROTECTED]>
>>>> To: "DBMail mailinglist" <[email protected]>
>>>> Sent: Wednesday, July 04, 2007 8:14 AM
>>>> Subject: Re: [Dbmail] Bug #579
>>>>
>>>>
>>>>> Jonathan Feally wrote:
>>>>>> What OS, kernel, glib, and db versions are you running? I tested
>>>>>> the search command by hand using the literal method and did not
>>>>>> have any problems at all. Also - your log file doesn't seem to
>>>>>> indicate anything about the child process crashing out, but rather
>>>>>> that the main parent process was told to shutdown down.
>>>>>
>>>>> Damn, I didn't even see that! You're right; the process is shutdown
>>>>> clean.
>>>>>
>>>>> Jul 2 23:23:15 lira dbmail/imap4d[6018]: Debug:[sql]
>>>>> dbmysql.c,db_query(+291): query [SELECT m.message_idnr,k.messageblk
>>>>> FROM dbmail_messageblks k JOIN dbmail_physmessage p ON k
>>>>> .physmessage_id = p.id JOIN dbmail_messages m ON p.id =
>>>>> m.physmessage_id WHERE mailbox_idnr = 507 AND status IN (0,1 ) AND
>>>>> k.is_header = '0' GROUP BY m.message_idnr,k.messageblk
>>>>> HAVING k.messageblk LIKE '%aa%']
>>>>>
>>>>> Jul 2 23:24:30 lira dbmail/imap4d[6000]: Debug:[server]
>>>>> server.c,server_run(+291): MainSigHandler(): got signal [15]
>>>>> Jul 2 23:24:30 lira dbmail/imap4d[6001]: Message:[server]
>>>>> pool.c,manage_stop_children(+470): General stop requested. Killing
>>>>> children...
>>>>>
>>>>>
>>>>> So that query is taking rather long, after which something/someone
>>>>> requested a general shutdown of the dbmail-imapd processes.
>>>>>
>>>>> There is no indication of a crash, but something is not right.
>>>>>
>>>>> --
>>>>> ________________________________________________________________
>>>>> Paul Stevens paul at nfg.nl
>>>>> NET FACILITIES GROUP GPG/PGP: 1024D/11F8CD31
>>>>> The Netherlands________________________________http://www.nfg.nl
>>>>> _______________________________________________
>>>>> DBmail mailing list
>>>>> [email protected]
>>>>> https://mailman.fastxs.nl/mailman/listinfo/dbmail
>>>>>
>>>>
>>>> _______________________________________________
>>>> DBmail mailing list
>>>> [email protected]
>>>> https://mailman.fastxs.nl/mailman/listinfo/dbmail
>>>>
>>>>
>>>
>>>
>>>
>>> _______________________________________________
>>> DBmail mailing list
>>> [email protected]
>>> https://mailman.fastxs.nl/mailman/listinfo/dbmail
>>>
>> _______________________________________________
>> DBmail mailing list
>> [email protected]
>> https://mailman.fastxs.nl/mailman/listinfo/dbmail
>>
>>
>
>
>
> _______________________________________________
> DBmail mailing list
> [email protected]
> https://mailman.fastxs.nl/mailman/listinfo/dbmail
>
--
_______________________________________________
DBmail mailing list
[email protected]
https://mailman.fastxs.nl/mailman/listinfo/dbmail
