Elasticsearch monitoring plugin check_es_system 1.11.1 released: Two bug fixes

Written by - 0 comments

Published on - Listed in Elasticsearch Monitoring ELK


A new version of check_es_system, an open source monitoring plugin to monitor Elasticsearch nodes and clusters, is available!

Version 1.11.1 is a bugfix release and fixes two important bugs which went under the radar.

Improper authentication handling

The first bug was reported in #38 and involves improper authentication handling. Without going into too much details, the check types readonly, master and tps were not properly checking when the authentication was invalid (typically a 401 unauthenticated error).

The other check types used the internal function getstatus which made sure that authentication was correct.

To solve this, all check types now use the internal function getstatus.

Thanks to Dan Johansson for reporting this and even providing a code fix!

Invalid address used

This bug has facepalm potential. When the Elasticsearch address (given by -H parameter) is an "archived" Elasticsearch instance at Elastic.co, the domain still responds to requests with a HTTP 200 status, however with a response of {"ok":false,"message":"Unknown resource."}.

The plugin interpreted this as everything is good and returned an OK return code without any output instead of failing.

$ ./check_es_system.sh -H outdated-cluster.cloud.es.io -P 9243 -S -u user -p pass -t status
$ echo $?
0

Something similar also happened when just a normal domain was used where no Elasticsearch at all is running; the plugin just returned OK without any output (although the output contained parse errors from the json parser):

$ ./check_es_system.sh -H www.claudiokuenzler.com -P 443 -S -u user -p secret -t status
parse error: Invalid numeric literal at line 1, column 10
parse error: Invalid numeric literal at line 1, column 10
parse error: Invalid numeric literal at line 1, column 10
parse error: Invalid numeric literal at line 1, column 10
parse error: Invalid numeric literal at line 1, column 10
parse error: Invalid numeric literal at line 1, column 10
parse error: Invalid numeric literal at line 1, column 10
parse error: Invalid numeric literal at line 1, column 10
parse error: Invalid numeric literal at line 1, column 10

$ echo $?
0

Both bugs are now fixed in version 1.11.1.

$ ./check_es_system.sh -H www.claudiokuenzler.com -P 443 -S -u user -p secret -t status
ES SYSTEM CRITICAL - Elasticsearch not available at this address www.claudiokuenzler.com:443



Add a comment

Show form to leave a comment

Comments (newest first)

No comments yet.