Processor Vs Process Monitoring
I opened the Monitor server ->Add data source -->Solaris (System (rexec:mpstat). And gave the userid, password, host name and port and connected to the server (in this case a processor). I can see Interrupts per sec, Interupts as threads etc.,.
I have a problem here. Under the processor we have several processes.
For example, we have LDAP process and one navigation process (for navigation process, we call Navigation server here).
In our tcpip based application, when the user signs in , the authentication is performed by Ldap.
If Ldap authentication is valid, then it requests the Navigation server to paint the appropriate page or pages.
At present in silk performer we can view only to the level of processor under solaris.
We can ses four processors as max.
When we conduct load test, how can we isolate at process level if there is a bottle neck in any one of these processes especially when the processor have around 10 processes.
Tip: We can type the following commands to get the info.
My question is why Silk performer hasnít provided this info through Graphics?
Information about a single process
The interesting information about a process is its consumption of resources. The
two main resources are memory and processor time.
The Virtual memory used by a process "Processname", in kilobytes. This eight 'x'
are defining the width of the field, so that we can easily separate the size and the
$ ps -e -o vsz=xxxxxxxx -o comm | egrep '^. . . . . . . . processname$' | cut -c -8
For the process usage (in percent), we use a similar command line:
$ ps -e -o pcpu=xxxxx -o comm | egrep '^. . . . . processname$' | cut -c -5
The number of dots and x's has to be the same and at least the same as the expected number of characters.
But I don't want to do it since we want to see this info graphically.
Do I have a solution in Silk Performer 4 or 4.5 for this.