How to measure the performance of ELK system?

Dear all,

I just started to learn and experiment with ELK in a logging related
project at work.

The answer I am looking for is that: say I am shipping tons of data into
logstash, do some basic transformation, and feed the data into
elasticsearch, finally view dashboard in Kibana. How long this whole
process take under load pressure?

I know this is a very general problem statement and the answer is like to
vary depending on machines, network etc. But I want to find out a generic
way to measure this, then I can tweak different machine/network
configurations.

What I have in mind is that:

  1. record the timestamp when generating the log entry, say t1
  2. add another timestamp in logstash, say t2
  3. somehow figure out a third timestamp t3, which is the time when the log
    entry is available in Kibana.

Then I will monitor the difference among t1, t2, and t3 in Kibana(assuming
the time on all machines are in sync), while scaling up the log(load)
generator.

Does this sound like a viable approach? if so, can you help me with some
instructions on how to get t2 and t3? (I have been struggling with this
step)

Thank you very much.

Yiming

--
You received this message because you are subscribed to the Google Groups "elasticsearch" group.
To unsubscribe from this group and stop receiving emails from it, send an email to elasticsearch+unsubscribe@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/elasticsearch/00663a14-835d-4724-b9c4-2b4c0a36d147%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

That seems like a sane idea, however adding the 3rd timestamp might be hard.
It'd probably be easier to let ES generate @timestamp and then leverage
that as your end point timestamp, as KB is simply pulling the data from ES.

On 29 January 2015 at 14:19, Yiming Li nicneo925@gmail.com wrote:

Dear all,

I just started to learn and experiment with ELK in a logging related
project at work.

The answer I am looking for is that: say I am shipping tons of data into
logstash, do some basic transformation, and feed the data into
elasticsearch, finally view dashboard in Kibana. How long this whole
process take under load pressure?

I know this is a very general problem statement and the answer is like to
vary depending on machines, network etc. But I want to find out a generic
way to measure this, then I can tweak different machine/network
configurations.

What I have in mind is that:

  1. record the timestamp when generating the log entry, say t1
  2. add another timestamp in logstash, say t2
  3. somehow figure out a third timestamp t3, which is the time when the log
    entry is available in Kibana.

Then I will monitor the difference among t1, t2, and t3 in Kibana(assuming
the time on all machines are in sync), while scaling up the log(load)
generator.

Does this sound like a viable approach? if so, can you help me with some
instructions on how to get t2 and t3? (I have been struggling with this
step)

Thank you very much.

Yiming

--
You received this message because you are subscribed to the Google Groups
"elasticsearch" group.
To unsubscribe from this group and stop receiving emails from it, send an
email to elasticsearch+unsubscribe@googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/elasticsearch/00663a14-835d-4724-b9c4-2b4c0a36d147%40googlegroups.com
https://groups.google.com/d/msgid/elasticsearch/00663a14-835d-4724-b9c4-2b4c0a36d147%40googlegroups.com?utm_medium=email&utm_source=footer
.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "elasticsearch" group.
To unsubscribe from this group and stop receiving emails from it, send an email to elasticsearch+unsubscribe@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/elasticsearch/CAEYi1X9N1oV-LD_sZOtfy538uy7VQBC4GhnEChLF3U5LRN-4QA%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.

Hi Mark,

Thanks for the reply. I will definitely try adding timestamp in the
mapping, as discussed
here: http://www.elasticsearch.org/guide/en/elasticsearch/reference/current/mapping-timestamp-field.html.

It seems that logstash will also generate a default @timestamp, if there is
no timestamp in the log itself, I will leverage that a too.

I will also look at the bigdesk plugin to identify elastic search system to
see at what moment resources are exhausted. Hopefully this combination can
give us some interesting result.

Yiming

On Thursday, January 29, 2015 at 2:01:05 PM UTC-8, Mark Walkom wrote:

That seems like a sane idea, however adding the 3rd timestamp might be
hard.
It'd probably be easier to let ES generate @timestamp and then leverage
that as your end point timestamp, as KB is simply pulling the data from ES.

On 29 January 2015 at 14:19, Yiming Li <nicn...@gmail.com <javascript:>>
wrote:

Dear all,

I just started to learn and experiment with ELK in a logging related
project at work.

The answer I am looking for is that: say I am shipping tons of data into
logstash, do some basic transformation, and feed the data into
elasticsearch, finally view dashboard in Kibana. How long this whole
process take under load pressure?

I know this is a very general problem statement and the answer is like to
vary depending on machines, network etc. But I want to find out a generic
way to measure this, then I can tweak different machine/network
configurations.

What I have in mind is that:

  1. record the timestamp when generating the log entry, say t1
  2. add another timestamp in logstash, say t2
  3. somehow figure out a third timestamp t3, which is the time when the
    log entry is available in Kibana.

Then I will monitor the difference among t1, t2, and t3 in
Kibana(assuming the time on all machines are in sync), while scaling up the
log(load) generator.

Does this sound like a viable approach? if so, can you help me with some
instructions on how to get t2 and t3? (I have been struggling with this
step)

Thank you very much.

Yiming

--
You received this message because you are subscribed to the Google Groups
"elasticsearch" group.
To unsubscribe from this group and stop receiving emails from it, send an
email to elasticsearc...@googlegroups.com <javascript:>.
To view this discussion on the web visit
https://groups.google.com/d/msgid/elasticsearch/00663a14-835d-4724-b9c4-2b4c0a36d147%40googlegroups.com
https://groups.google.com/d/msgid/elasticsearch/00663a14-835d-4724-b9c4-2b4c0a36d147%40googlegroups.com?utm_medium=email&utm_source=footer
.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "elasticsearch" group.
To unsubscribe from this group and stop receiving emails from it, send an email to elasticsearch+unsubscribe@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/elasticsearch/b8531c77-5a60-4114-b243-1398f7bec2ee%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.