i want to be able to measure the health of my ES deployment by looking at
various key counters/metrics like throughput, latency, load, disk IOs, etc.
that way i have visibility on how things are progressing..
are there any recommendations or does ES have an API for the same?
If you want to look at how these metrics look like (without alerts, for
example), I've used BigDesk and I liked it:
It seems to be the popular choice.
On Thursday, July 5, 2012 4:21:01 AM UTC+3, T Vinod Gupta wrote:
i want to be able to measure the health of my ES deployment by looking at
various key counters/metrics like throughput, latency, load, disk IOs, etc.
that way i have visibility on how things are progressing..
are there any recommendations or does ES have an API for the same?
For what it's worth we use SPM for Elasticsearch a LOT in all our
Elasticsearch work/engagements with clients. Radu mentioned Alerts, and
SPM has them in the current release, though we are keeping them limited to
a few test accounts in the backend for now. I think BigDesk is handy for
"ad-hoc" looking at ES, while SPM is handy if you want to continuously
monitor ES (and not just ES), if you want to be alerted when there are
issues, if you want to be able to look at historical data, etc.
On Wednesday, July 4, 2012 9:21:01 PM UTC-4, T Vinod Gupta wrote:
i want to be able to measure the health of my ES deployment by looking at
various key counters/metrics like throughput, latency, load, disk IOs, etc.
that way i have visibility on how things are progressing..
are there any recommendations or does ES have an API for the same?
Hi Otis,
this looks like a really cool tool. i was excited and tried to install it.
somehow it didnt work for me..
when i installed and made the changes to elasticsearch.in.sh file,
elasticsearch won't start. it starts when i revert the changes made to that
file. after installing spm, i just added this line to
elasticsearch.in.shat the very end -
For what it's worth we use SPM for Elasticsearch a LOT in all our
Elasticsearch work/engagements with clients. Radu mentioned Alerts, and
SPM has them in the current release, though we are keeping them limited to
a few test accounts in the backend for now. I think BigDesk is handy for
"ad-hoc" looking at ES, while SPM is handy if you want to continuously
monitor ES (and not just ES), if you want to be alerted when there are
issues, if you want to be able to look at historical data, etc.
On Wednesday, July 4, 2012 9:21:01 PM UTC-4, T Vinod Gupta wrote:
i want to be able to measure the health of my ES deployment by looking at
various key counters/metrics like throughput, latency, load, disk IOs, etc.
that way i have visibility on how things are progressing..
are there any recommendations or does ES have an API for the same?
On Sunday, July 8, 2012 10:22:53 PM UTC-4, T Vinod Gupta wrote:
Hi Otis,
this looks like a really cool tool. i was excited and tried to install it.
somehow it didnt work for me..
when i installed and made the changes to elasticsearch.in.sh file,
elasticsearch won't start. it starts when i revert the changes made to that
file. after installing spm, i just added this line to elasticsearch.in.shat the very end -
For what it's worth we use SPM for Elasticsearch a LOT in all our
Elasticsearch work/engagements with clients. Radu mentioned Alerts, and
SPM has them in the current release, though we are keeping them limited to
a few test accounts in the backend for now. I think BigDesk is handy for
"ad-hoc" looking at ES, while SPM is handy if you want to continuously
monitor ES (and not just ES), if you want to be alerted when there are
issues, if you want to be able to look at historical data, etc.
On Wednesday, July 4, 2012 9:21:01 PM UTC-4, T Vinod Gupta wrote:
i want to be able to measure the health of my ES deployment by looking
at various key counters/metrics like throughput, latency, load, disk IOs,
etc. that way i have visibility on how things are progressing..
are there any recommendations or does ES have an API for the same?
I've been fighting with some problems on ES 0.18.x (production) and ES
0.19.x in my dev environment and came up with a home brew solution for
performance monitoring Elasticsearch:
I like BigDesk for a heads up view, but Graphite is king for being able
compare data sets..
On Thursday, July 5, 2012 3:21:01 AM UTC+2, T Vinod Gupta wrote:
i want to be able to measure the health of my ES deployment by looking at
various key counters/metrics like throughput, latency, load, disk IOs, etc.
that way i have visibility on how things are progressing..
are there any recommendations or does ES have an API for the same?
speaking about bigdesk Otis is right, people can find it handy for quick
checking of the cluster (for now namely Node Stats) - that is what it was
build for. But for now it lacks features that would make it a solid
production tool like ability to store large amount of data into some
datastore, alerts and customizations. Personally, I am/was using bigdesk
especially when I was experimenting with elasticsearch as it allowed me to
quickly realize how to better plan resources for production deployment.
More then monitoring tool bigdesk is like an exploration tool.
You can also try to check elasticsearch-paramedic [1] by Karmi. But again,
IMO it lacks the same features like bigdesk that would make it a solid
production tool, though it brings some cool charts and some fancy signals
for red cluster has been implemented there recently!
I've been fighting with some problems on ES 0.18.x (production) and ES
0.19.x in my dev environment and came up with a home brew solution for
performance monitoring Elasticsearch:
I like BigDesk for a heads up view, but Graphite is king for being able
compare data sets..
On Thursday, July 5, 2012 3:21:01 AM UTC+2, T Vinod Gupta wrote:
i want to be able to measure the health of my ES deployment by looking at
various key counters/metrics like throughput, latency, load, disk IOs, etc.
that way i have visibility on how things are progressing..
are there any recommendations or does ES have an API for the same?
speaking about bigdesk Otis is right, people can find it handy for quick
checking of the cluster (for now namely Node Stats) - that is what it was
build for. But for now it lacks features that would make it a solid
production tool like ability to store large amount of data into some
datastore, alerts and customizations. Personally, I am/was using bigdesk
especially when I was experimenting with elasticsearch as it allowed me to
quickly realize how to better plan resources for production deployment.
More then monitoring tool bigdesk is like an exploration tool.
You can also try to check elasticsearch-paramedic [1] by Karmi. But again,
IMO it lacks the same features like bigdesk that would make it a solid
production tool, though it brings some cool charts and some fancy signals
for red cluster has been implemented there recently!
I've been fighting with some problems on ES 0.18.x (production) and ES
0.19.x in my dev environment and came up with a home brew solution for
performance monitoring Elasticsearch:
I like BigDesk for a heads up view, but Graphite is king for being able
compare data sets..
On Thursday, July 5, 2012 3:21:01 AM UTC+2, T Vinod Gupta wrote:
i want to be able to measure the health of my ES deployment by looking
at various key counters/metrics like throughput, latency, load, disk IOs,
etc. that way i have visibility on how things are progressing..
are there any recommendations or does ES have an API for the same?
Apache, Apache Lucene, Apache Hadoop, Hadoop, HDFS and the yellow elephant
logo are trademarks of the
Apache Software Foundation
in the United States and/or other countries.
