Hi guys!
Well.We have RM 8.1.502.80 under Windows 2008r2 x64 and MCP 8.1.504.94 under Windows 2003Sr2 x86
And right now I’m open another service request against MCP. As i see - it’s again problem with memory leak in MCP or RM , or in both them. Now it’s the 4(FOURTH) SR for the last quarter with the similar symptoms - RM or MCP stop to respond on SIP request. Then i look in their log - i just don’t see any invite from SIP servers/client. But i’m always saw an not normal memory grow for “silent” GVP service. So that i do next ? Collect all log from all service for a last 2 day. Check the network access. Create service dump. Then reboot the “silent” serice and voila - service start to respond again.
Then i send all collected info to Genesys and ? Yes i have again everything but not the answer on my question:
You have a problem with SIP server" WTF?! Yes. And we correct it in next year …
You have a problem with network - check it
You have a problem with something else - check it
You have a problem…
DAMN! I know that I HAVE THE PROBLEM! 4(FOURTH!!!) serivice request with similar symptoms - but in MCP and RM: “everything OK”.
So..It’s not a kind of “cry about whole world” as you can think =) I realy do expect that’s what:
If you have…in production, in you lab envirovment, in production or lab envirovment of you colleagues, of you friends or friends colleagues, - altogether if you have the similar problem - share the SR number what you open in Genesys TS. Just send it details in my personal e-mail.
Thank to all. Hope it’s can help to all comunity.
WBR Tim
I have just finished my implementation and now concerned on what I will expect…lol
Just dummy question, you are running last version of everything, right?
Which is your call volume?
Well, it’s not a dammy - for Genesys it’s a right question! For all software except GVP =( Coz last time then i change MCP GA to higer major release(that i maid coz it’s need to ZOOM) i one month slept on work before i get the HF (our current release). So right now i have the latest GA for RM and latest “stable” HF for MCP.
If we look in release notes - we don’t found any of such symptoms and issue solved.
WBR
Yeah I know that maybe on release notes it might not be pointed…but you know, I had sometimes weird situations also not found on release notes and an upgrade by magic solved the issue…so who knows… ;D
Hi Timur,
we are registering problems with new GVP solution(RM+MCP) too,but I am not sure if same as you. When you write me more specific desciption,I can compare it with our issues and next share commons SR. I have to say,that we have registered a lot of issues at the begin of deployment new GVP8.1,but most of these issues has been solved. However.currently I found “big” bug on specific deployment type of Genesys SIP + MSML(ie.GVP8).
I highly recommended do not use TCP as transport protocol for GVP with SIP server in HA mode
I know this really doesn’t help this thread, but for those contemplating deploying GVP I would recommend using Linux. We have deployed GVP 8.1.5 in HA setting with 3 MCP’s (No ASR/TTS) on Red Hat Linux 5 64-bit and have not had any failures for 3 months in production environment - handling average of 65k sessions per day, including MCP for CPA. From what I’ve seen, most problems and hot fixes appear to target the Windows platform. I’m not a Windows hater, far from it, but the solution does seem more stable on Linux.
One of many reasons for memory leak is related to local (non US) characters in SIP messages. Usually it is forgotten but any non UTF-8 characters send/received in headers to/from SIPS in long term are causing MCP crashes. Premanent solution to this problem is to set properly encoding option on SIPS side.
