Письмо #20504 Списка Рассылки CGatePro@list.communigate.ru
От Кого: Nicolas Hatier nicolas.hatier@niversoft.com <CGatePro@list.communigate.ru>
Кому: CommuniGate Pro Russian Discussions <CGatePro@list.communigate.ru>
Тема: Re: [CGP] [*] CommuniGate Pro 6.3.11
Дата: Wed, 11 May 2022 23:00:32 -0400

I confirm it's very likely linked to either the event or the trigger, as it happened again, this time on a test Windows server, immediately after the SendURL call was completed.


On 2022-05-11 07:34, Nicolas Hatier nicolas.hatier@niversoft.com wrote:

Sorry, core dumps were not enabled.

I don't use the MB4 mailbox format

I don't use RPOP on that server.

On the last crash, CGP crashed again after a few seconds when I restarted it. Since the last log line was the SendURL trigger, I renamed Settings/EventsHandlers.settings and CGP started successfully. Once the queue cleared, I put back the file in place and restarted CGP again, and it started OK.

So, this might be related to either the numQueuedMessages event or the SendURL trigger, or to something that happens just after the trigger.



On 2022-05-11 01:12, Dmitry Akindinov dimak@communigate.ru wrote:
Hello,

On 2022-05-11 00:04, Nicolas Hatier nicolas.hatier@niversoft.com wrote:

Since upgrading to 6.3.11, I got a three random crashes. One occurred two weeks ago, and two today.

Previously I was running 6.3.3 and it was very stable for me, as all previous versions.

I have this in /var/log/maillog for today's crashes.

May 10 13:03:08 server CommuniGatePro[2866998]: Exception signal: 11, trying to recover
May 10 13:03:08 server CommuniGatePro[2866998]: Exception in <SysThread 2F13F00 Main>
May 10 16:29:20 server CommuniGatePro[3235153]: Exception signal: 11, trying to recover
May 10 16:29:20 server CommuniGatePro[3235153]: Exception in <SysThread 31A6F00 Main>

It would help to get a core dump from these crashes. Were any created?

There was nothing relevant in CGP logs for the first crash two weeks ago. Today's crashes were both just after a numQueuedMessages event was triggered and the associated SendUrl trigger handler ran successfully. This may, however, be a red herring.

The crashes happened in the Main thread, which is mostly used to run periodic Chronos-driven tasks like mailbox archivung or RPOP/RIMAP polling. Do you ise any of those?
Do you use the new MB4 mailbox format?

Running CGP on CentOS 8.4.2105

*Nicolas Hatier, ing.* <nicolas.hatier@niversoft.com>
Niversoft idées logicielles - http://www.niversoft.com




On 2022-04-26 08:05, Mailing List Admin CGPUpdate@list.communigate.ru wrote:
Выпуск версии с исправлениями

Доступен CommuniGate Pro 6.3.11.
Это обновление настоятельно рекомендовано к установке.

6.3.10 26-апр-2022

* Samoware: в состав входит Samoware HTML версии 6.3.11.
* MAPI: в состав входит MAPI Connector версии 1.54.12.39.
* Исправление: SIP: 6.3.8: обработка URI со специальными суффиксами
могла приводить к аварийной остановке.
* Исправление: AIRSYNC: 6.3c1: обновление события с приложениями могло
приводить к аварийной остановке.
* Исправление: CLI: 6.1, 6.3c1: некоторые методы не проверяли наличие
административных прав.

Полная история разработки:
http://www.communigate.ru/CommuniGatePro/russian/History63.html

Release notes in English:
http://www.communigate.ru/CommuniGatePro/History63.html





Подписаться (Прямо) Подписаться (Дайджест) Подписаться (Оглавление) Отписаться Написать Listmaster-у