rstatd for UNIX Resource Monitoring


| posted in , , , , | 0 comments




Monitoring UNIX based Servers with LoadRunner is extensively in use with rstatd daemon.

How to check the rstatd is running :

Use the commands "rpcinfo -p" and "svcs -a"
and check whether you can connect to port that is listening rstatd daemon via rup/rusers/telnet(Not sure , it may vary)

What we can monitor :

Here are the list of parameters we can monitor with rstatd daemon in UNIX:

Problems faced in real time with rstad by a loadtester for example:
It is posted on 22nd August and there is no update for this post in that unix forum

We have an admin who is running Mecury LoadRunner. One part of it generates
load to our Solaris 10 Web and Application servers (SPARC), and the other
collects
performance statistics via the Solaris RPC "rstatd" interface.

Now, randomly, and independent of the load placed on the Solaris Web/App
servers (it
doesn't matter), the Windows LoadRunner program that connects to those
servers
to collect performance data, via the rstatd interface, dies generating the
following
error message (I believe performance collector program is written in Java):

Monitor name :UNIX Resources. Internal rpc error (error code:2).
Hint: Check that RPC on this machine is up and running. Check that rstat
daemon
on this machine is up and running (use rpcinfo utility for this
verification).
Details: RPC: RPC call failed.
RPC-TCP: recv()/recvfrom() failed.
RPC-TCP: Timeout reached. (entry point: Factory::CollectData). [MsgId:
MMSG-47197]

Now "rpcinfo -p" and "svcs -a" show that the rstatd daemon is running, and I
can
connect to it via rup/rusers/telnet to the port it is listening on (this has
been the
case along - this has not changed).

This just started happening when we moved the Solaris web/app machines to
another
network (but all that was changed were their IP addresses, and nothing
else). The load
generator servers themselves as well as the LoadRunner rstat performance
data collector
servers were not moved to other networks.

We think its network related but are not sure - it could be the LoadRunner
rstat data
collector program itself (written probably in java and running a PC) having
timeout
issues.

I've done tcpdumps and looked ... and the rstat calls look totally normal.

Anyone seen this / have advice?

TIA,
Noel

If anybody knows , please reply and help the beginners to learn.

Comments :

0 comments to “ rstatd for UNIX Resource Monitoring ”

Post a Comment