busindre / elasticsearch-cacti Goto Github PK

Hi,

First of all want to thank you about great work.

Found some problem not sure what it is related for (cacti itself or importing or maybe just mine version or maybe some shielding). So just mentioning it here for anyone who will got the same.

Mine cacti version 1.2.2

Problem that graphs for Index not working at all and I don't see any errors just result that data input method "Elasticsearch index" didn't work at all as result no rrds and no graphs :)

Data Collection->Data Input Methods->Elasticsearch index (Edit)
And change "Input String" from:
/bin/bash <path_cacti>/scripts/elasticsearch.sh <hostname> index "<index_name>"
to this:
/bin/bash <path_cacti>/scripts/elasticsearch.sh <hostname> index <index_name>

Cleared poller cache and it started to work fine.

Suppose maybe apply some shielding it will work but actually for me without quotes is enough :)

thanks

I'm trying out the cacti templates instead of the pain that appears to be marvel/kibana itself but have a small issue.
Running the elasticsearch.sh script manually, gives me results from the elasticsearch nodes/cluster.
Running (the recompiled) spine manually with the specific hostid, also gives me results.
However, allowing cacti itself to run spine, gives me back "output:U" and no data.

examples
09/09/2016 03:59:41 PM - SPINE: Poller[0] Host[93] TH[1] DS[3682] SCRIPT: /bin/bash /usr/share/cacti/scripts/elasticsearch.sh elasticsearch01 cluster, output: active_primary_shards:53 active_shards:106 relocating_shards:0 initializing_shards:0 unassigned_shards:0 delayed_unassigned_shards:0 cluster_red:0 cluster_yellow:0 cluster_green:1 immediate:0 urgent:0 high:0 normal:0 low:0 languid:0

09/09/2016 04:00:10 PM - SPINE: Poller[0] Host[93] TH[1] DS[3683] SCRIPT: /bin/bash /usr/share/cacti/scripts/elasticsearch.sh elasticsearch01 cluster, output: U

any ideas why this would be?
Thanks,
Sean

hi,
small issue in datasource

old_used_in_bytes shows nan
problem that it's equal to 100 and sure old memory in java far more bigger value.

Templates->Data Source->Elasticsearch JVM (Edit)->old_used_in_bytes and change "Maximum Value ("U" for No Maximum)" from 100 to U

It's better to do before creating charts cause rrd you need to update manually or delete them to apply changes

Anyway there is heap chart to estimate old memory size