Server crash every 1-2 days after updating to new Chromium

[lokiys]

Updated zenno to 7.6 and as well I was forced to update to Chromium, as CEF does not worked anymore for my tasks and now two different servers crash by not known issues. I have to connect with remote access control tools to hard restart them. Basically after 24h run there is no possible to login sever, and all you can see is black screen. I have opened a support ticket but it does not seem there is somebody interested in this issue. Two month and there is no any news from support in tickets and how this will be fixed. @VladZen can you look maybe?
Conversation ID:37628

 

[VladZen]

I'll review the ticket.
For now, I would suggest you trying version 7.7.0.0. Perhaps, it will work more stable.
Chromium is still beta in both 7.6 and 7.7, but 7.7 includes latest optimizations.

 

[VladZen]

Reviewed the tickets...
The only problem by logs is that browser renderers crashes by not enough memory.
Application: instance_cr_helper_sta.exe
Framework Version: v4.0.30319
Description: The process was terminated due to an unhandled exception.
Exception Info: System.OutOfMemoryException
at Global.ZyanHelpers.ZyanErrorHandler..ctor(Int32)
at ChromiumInstance.Renderer.Program.Main(System.String[])
at ChromiumInstance.Renderer.Sta.Program.Main(System.String[])

There is no even windows system dump after server crash.

Is this option enabled in zennoposter settings?

 

[lokiys]

Reviewed the tickets...
The only problem by logs is that browser renderers crashes by not enough memory.
Application: instance_cr_helper_sta.exe
Framework Version: v4.0.30319
Description: The process was terminated due to an unhandled exception.
Exception Info: System.OutOfMemoryException
at Global.ZyanHelpers.ZyanErrorHandler..ctor(Int32)
at ChromiumInstance.Renderer.Program.Main(System.String[])
at ChromiumInstance.Renderer.Sta.Program.Main(System.String[])

There is no even windows system dump after server crash.

Is this option enabled in zennoposter settings?
Посмотреть вложение 89120

On one server it was enabled on other not. I do not remember from what server there is report posted in tickets.

BUT

This is not the issue I'm reporting. Everything is in tickets, server gets black screen and it is not possible to login when bots have worked 1-2 days. I guess instance crash can not crash the whole server right ? + I have moved all bots to servers You have suggested me in tickets, all is Intel and it works with like 10 threads. If there is no any cache or any other memory leak bug it should work with 10 threads very good on such server right ?
Server in question: Dedicated Root Server Hosting - Hetzner Online GmbH

 

[VladZen]

On one server it was enabled on other not. I do not remember from what server there is report posted in tickets.

BUT

This is not the issue I'm reporting. Everything is in tickets, server gets black screen and it is not possible to login when bots have worked 1-2 days. I guess instance crash can not crash the whole server right ? + I have moved all bots to servers You have suggested me in tickets, all is Intel and it works with like 10 threads. If there is no any cache or any other memory leak bug it should work with 10 threads very good on such server right ?
Server in question: Dedicated Root Server Hosting - Hetzner Online GmbH

After reviewing the logs, developers strongly advise you to enable resource monitor on both servers and see how it works.

 

[lokiys]

After reviewing the logs, developers strongly advise you to enable resource monitor on both servers and see how it works.

It was enabled on one server and there is/was no any difference how they work.

 

[VladZen]

This is not the issue I'm reporting. Everything is in tickets, server gets black screen and it is not possible to login when bots have worked 1-2 days. I guess instance crash can not crash the whole server right ? + I have moved all bots to servers You have suggested me in tickets, all is Intel and it works with like 10 threads. If there is no any cache or any other memory leak bug it should work with 10 threads very good on such server right ?
Server in question: Dedicated Root Server Hosting - Hetzner Online GmbH

It does no matter, server crashes or goes down...
According to logs, instance render process crashing causes memory leak which can be the reason for server overload. Therefore we recommend to enable resource monitor option.
Have you tried suggested 7.7 version? How does it work?
If server still goes down, cloud you please make fresh report from this version?
Also, do you use profile-folders with Chromium?

 

[lokiys]

It does no matter, server crashes or goes down...
According to logs, instance render process crashing causes memory leak which can be the reason for server overload. Therefore we recommend to enable resource monitor option.
Have you tried suggested 7.7 version? How does it work?
If server still goes down, cloud you please make fresh report from this version?
Also, do you use profile-folders with Chromium?

I have installed v7 as soon as it came out as test version, but experienced some problems and downgraded, so, please release version and I will test again.

Again I told resource monitor was enabled for one server, and there were no any differences, so no help from that.

I do not manage profiles by myself, so it means it works default as is.