Dear Cindy,

Will it be possible to get answers for these questions or I'm just asking for too much? :)

Dear Cindy,

Thanks a lot for your help!

Dear Anthony,

Since new version is not available for us for now, I think the best for us then would be to try 2.1.3 version in order to see it in practice, after that I'll ask you particular questions

Should it work on CentOS 6?

Also, is it possible to get some information about Zenoss HPC usage on RoadRunner? Or this info is not publicly available?

UPD: Our admin tried to install 2.1.3 (got it here: http://dev.zenoss.org/trac/browser/branches/zenoss-2.1.3-hpc/RPMS) on CentOS release 6.3 (Final) and the following errors apperared:

"

Error: Package: zenoss-2.1.3-1.hpc.el5.x86_64 (/zenoss-2.1.3-1.hpc.el5.x86_64)

           Requires: libmysqlclient.so.15()(64bit)

Error: Package: zenoss-2.1.3-1.hpc.el5.x86_64 (/zenoss-2.1.3-1.hpc.el5.x86_64)

           Requires: libpq.so.4()(64bit)

Error: Package: zenoss-2.1.3-1.hpc.el5.x86_64 (/zenoss-2.1.3-1.hpc.el5.x86_64)

           Requires: libgnutls.so.13(GNUTLS_1_3)(64bit)

Error: Package: zenoss-2.1.3-1.hpc.el5.x86_64 (/zenoss-2.1.3-1.hpc.el5.x86_64)

           Requires: libmysqlclient.so.15(libmysqlclient_15)(64bit)

Error: Package: zenoss-2.1.3-1.hpc.el5.x86_64 (/zenoss-2.1.3-1.hpc.el5.x86_64)

           Requires: libmysqlclient_r.so.15(libmysqlclient_15)(64bit)

Error: Package: zenoss-2.1.3-1.hpc.el5.x86_64 (/zenoss-2.1.3-1.hpc.el5.x86_64)

           Requires: libgnutls.so.13()(64bit)

Error: Package: zenoss-2.1.3-1.hpc.el5.x86_64 (/zenoss-2.1.3-1.hpc.el5.x86_64)

           Requires: libmysqlclient_r.so.15()(64bit)

"

Maybe you can tell us how this could be fixed?

Hi Laura,

Concerning RoadRunner there are several questions that are interesting for us:

• What types of data are collected and processed by Zenoss HPC and how frequently? For example, CPU usage - once per minute, nodes network activity - every 10 seconds, room temperature - once per 5 minutes etc.
• How many instances of Zenoss HPC are needed for monitoring RoadRunner? What hardware is used for that purposes?
• As well as we understand, Zenoss HPC is using a set of rules for issue tracking. Accordingly at some point (new sensor value came, timer triggered) these rules are being checked. How many such rules are you using (concerning RoadRunner), and how often this checking is made?
• Do you consider Zenoss HPC as sufficient for RoadRunner? Do you use any other monitoring systems?

Hope that would not be too much  to ask

Thanks in advance!

UPD: We've tried to build our own rpm, but it seems that spec file is designed for other directory layout...

The following error occurs:

<...>

Executing(%build): /bin/sh -e /var/tmp/rpm-tmp.P5JpNw

+ umask 022

+ cd /home/vadim/rpmbuild/BUILD

+ cd zenoss-2.1.3

+ LANG=C

+ export LANG

+ unset DISPLAY

+ rm -rf /home/vadim/rpmbuild/BUILDROOT/zenoss-2.1.3-1.hpc.el5.x86_64

+ export ZENHOME=/opt/zenoss

+ ZENHOME=/opt/zenoss

+ export ZOPEPASSWORD=zenoss

+ ZOPEPASSWORD=zenoss

+ export PYTHON=/usr/bin/python

+ PYTHON=/usr/bin/python

+ export DESTDIR=/home/vadim/rpmbuild/BUILDROOT/zenoss-2.1.3-1.hpc.el5.x86_64

+ DESTDIR=/home/vadim/rpmbuild/BUILDROOT/zenoss-2.1.3-1.hpc.el5.x86_64

+ ZEN_BUILD_DIR=/home/vadim/rpmbuild/BUILD/zenoss-2.1.3

+ . /home/vadim/rpmbuild/BUILD/zenoss-2.1.3/build-functions.sh

/var/tmp/rpm-tmp.P5JpNw: line 54: /home/vadim/rpmbuild/BUILD/zenoss-2.1.3/build-functions.sh: No such file or directory

error: Bad exit status from /var/tmp/rpm-tmp.P5JpNw (%build)

The needed build_functions.sh is located at different directory (/home/vadim/rpmbuild/BUILD/zenoss-2.1.3/inst/build-functions.sh).

We've tried to change spec file a bit, but another errors occur...

Is it possible for you to tell us what should be changed in this file (or maybe some preliminary settings should be made)?

First of all, thank you for the answers!

"I will try to answer your questions concerning Roadrunner and see if I can get one of our sys admins to look at you rpm errors."

Any help about rpm issue will be great!

"We collect ping up/ping down, cpu usage (at the node level) using our scheduler, network data from syslog, cpu temperature, correctable and uncorrectable memory errors from syslog."

And how often do you collect such data? Once per second?

Thanks again!

Vadim