Hi,
I Have te following problem, I use PerfCounter = uEplPoolPagedMemory,"\Proces(Epl)\Niet-wisselbaar geheugen: bytes",10 in my client conf file on a Windows 2003 server. I use this to read the non-paged memory pool usage of the Epl process (on a server with dutch language).
I can get the value of this no problem, however there are multiple items named Epl, which I have to address as ...\Proces(Epl#1)\... ...\Proces(Epl#2)\... etc to read the data. Since there are over 40 processes that are named this way, I would have 40 rows in my conf file. Not very neat and when the 41st epl is started that one is not read obviously.
What I would like is to create some way to get the value of the epl process with the highest non-paged memory pool usage. The reason for this is that the software, running on this server, is highly unstable. The only warning is the heavy usage of non-paged memory by one of the epl processes just prior to a server crash. I would like to get ahead of that, so that I can kill the process manually and stop the server from melting
Actually it's quite like proc_info[epl,vmsize,max] but that reads the virtual memory used in stead of the non-paged memory pool.
Anyone a bright idea?
I use version 1.4.5 btw, both server and client.
I Have te following problem, I use PerfCounter = uEplPoolPagedMemory,"\Proces(Epl)\Niet-wisselbaar geheugen: bytes",10 in my client conf file on a Windows 2003 server. I use this to read the non-paged memory pool usage of the Epl process (on a server with dutch language).
I can get the value of this no problem, however there are multiple items named Epl, which I have to address as ...\Proces(Epl#1)\... ...\Proces(Epl#2)\... etc to read the data. Since there are over 40 processes that are named this way, I would have 40 rows in my conf file. Not very neat and when the 41st epl is started that one is not read obviously.
What I would like is to create some way to get the value of the epl process with the highest non-paged memory pool usage. The reason for this is that the software, running on this server, is highly unstable. The only warning is the heavy usage of non-paged memory by one of the epl processes just prior to a server crash. I would like to get ahead of that, so that I can kill the process manually and stop the server from melting

Actually it's quite like proc_info[epl,vmsize,max] but that reads the virtual memory used in stead of the non-paged memory pool.
Anyone a bright idea?
I use version 1.4.5 btw, both server and client.

Comment