What is Ganglia?
Posted by Matt Massie in FYI on November 9, 2005
Ganglia is a scalable distributed monitoring system for high-performance computing systems such as clusters and Grids. It is based on a hierarchical design targeted at federations of clusters. It leverages widely used technologies such as XML for data representation, XDR for compact, portable data transport, and RRDtool for data storage and visualization. It uses carefully engineered data structures and algorithms to achieve very low per-node overheads and high concurrency. The implementation is robust, has been ported to an extensive set of operating systems and processor architectures, and is currently in use on thousands of clusters around the world. It has been used to link clusters across university campuses and around the world and can scale to handle clusters with 2000 nodes.
Ganglia is a BSD-licensed open-source project that grew out of the University of California, Berkeley Millennium Project which was initially funded in large part by the National Partnership for Advanced Computational Infrastructure (NPACI) and National Science Foundation RI Award EIA-9802069. NPACI is funded by the National Science Foundation and strives to advance science by creating a ubiquitous, continuous, and pervasive national computational infrastructure: the Grid. Current support comes from Planet Lab: an open platform for developing, deploying, and accessing planetary-scale services.
Overlaying timeshifted data
Posted by vuksan in Announcements on May 16, 2012
Often times when troubleshooting an issue it is useful to compare current data to data from a previous period. For example you observe high load on one machine and want to compare it to the load from a previous day. In the past you would simply open up two browser windows with data from each period and compare them. This may prove tricky as you may be dealing with different Y scales. To help with that we have recently added ability to overlay timeshifted data onto any host metric graph in Ganglia. What this means is that next to each metric graph you will see a button that says Timeshift. Clicking on it will overlay data from the same period prior to it e.g. if you are viewing hour it will overlay data from an hour ago, if you are viewing day it will overlay data from a day ago etc. For example this shows one minute load average overlaid with data from yesterday
Timeshifted load one average
whereas graph below displays data from a week ago overlaid over this weeks data
Timeshift overlay CPU user
Preliminary support for this feature has already landed in the trunk of our Ganglia Web repository
https://github.com/ganglia/ganglia-web
It is full usable although we are working on refining it. This should be included in Ganglia Web 3.5.0 when it’s fully ready.
Ganglia Web 3.4.2 released
Posted by Bernard Li in Announcements, Releases on May 8, 2012
The Ganglia team is pleased to release Ganglia Web 3.4.2. Notable changes are:
- Improvements to the Live Dashboard
- Fixed the aggregate graphs metric auto complete which was broken in 3.4.1
- Add ability to specify critical and warning thresholds which can be used in Live Dashboard and Views
- Minor bug fixes
This release can be downloaded here.
Ganglia 3.3.7 released
Posted by Bernard Li in Announcements, Releases on May 8, 2012
The Ganglia team is pleased to announce the release of 3.3.7. The following issues have been fixed in gmond:
- BUG100: Fails to start in Solaris containers
- Fails to start when no address on the network interface (added retry_bind parameter)
- BUG321: Fails to start when Solaris CPU is in FAILED state (segmentation fault)
The release can be immediately downloaded here.
Ganglia 3.3.6 released
Posted by Bernard Li in Announcements, Releases on April 19, 2012
The Ganglia Development Team is happy to announce the release of Ganglia 3.3.6. This release fixes the following bug:
BUG327: memory leak when receive channel is not configured or not hearing any data
Ganglia 3.3.5 released
Posted by Bernard Li in Announcements, Releases on April 5, 2012
The Ganglia Development Team is happy to announce the release of Ganglia 3.3.5. This is mainly a bugfix release for 3.3.1 and all users of the 3.3.x series are encouraged to upgrade.
For more details regarding this release, please refer to the Release Notes. Downloads are available at SourceForge.
Work is already under way for the next release, please stay tuned for upcoming enhancements to the project!
Adding trendline to your graphs
Posted by vuksan in Uncategorized on March 8, 2012
As part of our upcoming release of Ganglia 3.3.2 you will have the ability to add a trend line to any host metric. Let’s say you wanted to know when you will need to add more disk space to a particular machine.
We are “safe” for about a year. Or let’s say you wanted to see how your web server 90th percentile has trended.
This takes calculates trend based on last six month of data (which is a default). That looks pretty good however let’s see how it looks if we take 12 months of data
Not so good. Trending interface has easy controls that allow you to easily change how far to extend the trend line and how far back to go
e.g.
Enjoy.
Ganglia 3.3.1 released
Posted by vuksan in Uncategorized on February 8, 2012
We are happy to announce release of Ganglia 3.3.1. This release provides several enhancements to our web frontend as well as additional capabilities for exporting data to Graphite.
Please refer to our release note. You can download the latest release at SourceForge.
Ganglia 3.3.0 released
Posted by vuksan in Uncategorized on February 1, 2012
We are happy to announce the release of Ganglia 3.3.0. Highlights of this release are
- Ganglia Monitor Core is now shipped with our second generation Ganglia Web UI.
- Gmetad Daemon now supports sending metrics to Graphite (carbon)
- sFlow supports additional metric sources such as JMX, memcache, Apache
- Gmond now comes bundled with a number of additional metrics e.g. disk statistics, network interface utilization etc.
Full Release notes can be found on Github.
You can download the release from SourceForge.
Please report any issues with the release to our GitHub issue tracker.
Putting metrics data on a common “bus”
Posted by vuksan in Uncategorized on December 29, 2011
Patrick Debois has kicked of an interesting set of projects to put metric information on a common “bus”. For example he has implemented a ruby based daemon that parses Ganglia gmond packets and puts them on a ZeroMQ pub/sub bus. Once it’s there you can “subscribe” with a client of your choice and do transforms to the data e.g.
- feed graphite or another monitoring tool
- insert data into a SQL database
- feed Nagios using passive checks
Thanks to Patrick for a great idea and implementation. Now let’s get to work on useful subscribers.
Gmond-zmq https://github.com/jedi4ever/gmond-zmq
Statsd-zmq https://github.com/jedi4ever/statsd
Ganglia Web 2.2.0 released
Posted by vuksan in Uncategorized on December 13, 2011
Today we released version 2.2.0 of the Ganglia Web interface. New features present in this release are described in our “Upcoming Web Features post“.
Release can be downloaded from https://sourceforge.net/projects/ganglia/files/gweb/2.2.0/
If you need help installation guide is here.
Polls
Loading ...