Hi Bruce, list,
I'm in the last stages of building a vmware monitoring extension for xymon that does not require any changes to the ESX-hosts monitored. It uses the vMA (more precisely the vCLI installed on the vMA) to fetch the information from the vmware hosts (see below for a list of features).
My apologies to the list for announcing both a not-yet-public-available and a "light-commercial" extension here.
This thing is named "Vmware-Monitor for Xymon (VMX)" and requires the following environment:
- vMA with hobbit- or xymon-client installed (hobbit-4.2.0 w/o all-in-one patch works fine)
- xymon-server with working user-channel and a working >=gawk-3.1.3
- in particular _no_ vCenter is needed (as a nice side-effect monitoring of vmware-server-2.0.x-hosts is possible too) The names of the VMs should be identical to their hostnames.
Currently the following features are implemented (short version):
- central configuration of VMware-monitoring on the XYmon server
- current CPU- and memory-usage, both of the ensemble of all hosts and for each individual VMware host and running VM (and graphical trending of these and more of course)
- per VM status-page with info about VMDKs, vCPUs, memory, ... and vmware-tools (along with additional infos if tools are running)
- overview and tracking of virtual machines on each VMware host, no matter if the VM was moved by vmotion, DRS, manually added/removed, ... So it's possible to easily determine when a VM was started or stopped on a particular host!
- warning if a host is in maintainance-mode or has the "reboot-needed" flag set
- vMEM/pMEM-ratio, i.e. how much of the physically available RAM is assigned to the running VMs
- one-click overview/summary for the following metrics: o number of monitored hosts (along with software version and patch-level) o number of running VMs o total CPU (MHz) available o CPU (MHz) used o total vCPU-count running o total physical RAM available o virtual RAM used (both "hostmemoryusage" and "guestmemoryusage")
- ressource friendly: low monitoring overhead due to non-forking design on the server-side and minimized disk-access to avoid I/O problems
- all status-pages are hyperlinked to each other so navigation from the overview-page down to a particular VM is quick and easy
The roadmap for possible additional features currently contains (no particular order):
- disk-IO
- network-IO
- duplicate VM detection
- configuration of alarm-triggers (e.g. for high cpu-/mem-usage)
Currently a ESX-cluster with 7 ESX 3.5 hosts is monitored without problems. ESX 4 / vSphere has been verified to work too as.
Documentation is not complete yet but installation instructions, explanation of the status columns generated and graphing-details and -options are already written.
This extension will be available at a very affordable price. Of course a portion of the revenues will go to Henrik for providing XYmon at all -- the "best monitoring solution"(tm).
If you are interested (either in getting an e-mail with the official announcement, in beta-testing or if you have questions) please let me know off-list.
A Happy New Year to you all,
ThomasOn Mon, 21 Dec 2009, White, Bruce wrote:
Hi all,
I saw a couple of threads on this topic from January and was wondering if there is any newer insight into this question. At that time, it appeared that many were just installing a RedHat based version of the Xymon (hobbit) client (with modifications to the df [vdf] and top [esxtop] commands) on their ESX servers despite the warning that it may void your support. I have started down the SNMP monitoring path and have found that most of the fields listed in VM's MIBs are not populated. So, are people still running modified versions of the hobbit client without issue? Has anyone made any real progress on monitoring ESX servers via SNMP?
Thanks in advance,
Bruce
Bruce White Senior Enterprise Systems Engineer | Phone: 630-671-5169 | Fax: 630-893-1648 | bewhite at fellowes.com | http://www.fellowes.com/
Disclaimer: The information contained in this message may be privileged and confidential and protected from disclosure. If the reader of this message is not the intended recipient or an employee or agent responsible for delivering this message to the intended recipient, you are hereby notified that any dissemination, distribution or copying of this communication is strictly prohibited. If you have received this communication in error, please notify us immediately by replying to the message and deleting it from your computer. Thank you. Fellowes, Inc.
| IT-Beratung Eckert | Hartkirchweg 54 | tel: 0761/ 594 9898 | Thomas Eckert | 79111 Freiburg i.Br. | fax: 0761/ 594 9899 | http://www.it-eckert.de/