add filter to curl request to prevent oom exceptions

[c-kr]

Sometimes the apis response gets very long because it logs every http requests by default. On our test machine the following grep cmd filled up the 24gb memory within seconds:

if [[ -n $user ]] || [[ -n $(echo $esstatus | grep -i authentication) ]] ; then

We were able to mitigate it by adding a filter to the api requests to include only those values which are needed. This probably needs some more testing before merging, we only tested the memory mode

[Napsty]

~~Lets give #42 a bit more time to rebase the PR, then adjust this PR and get a new version out.~~ Nevermind, we'll go ahead with this one first and then we can do #42. Very interesting that your Elasticsearch shows so much data though. Could it be a misconfigured ES at the cause? On which check type did you see the huge data in the API response?

[c-kr]

I am not an elastic guy but i don't think it is misconfigured, just heavily used which leads to logging of a lot open http connections in the api output. It only happens sometimes a day.

We see this huge data in all checks used by the getstatus() method with endpoint (/stats).

[Napsty]

Thanks for the info @c-kr

We see this huge data in all checks used by the getstatus() method with endpoint (/stats).

That is certainly strange. I will do some tests with my Elasticsearch instances and maybe I can reproduce this sometime. What kind of Elasticsearch did you see that on? Local self installed or a Cloud instance? Which version?

Nevertheless, I think your contribution makes sense and will be merged after tests. :+1:

[c-kr]

What kind of Elasticsearch did you see that on? Local self installed or a Cloud instance? Which version?

Self Installed Elasticsearch Version 7.17.1

Nevertheless, I think your contribution makes sense and will be merged after tests. 👍

Thanks

[Napsty]

disk check does not work anymore with that change.

Before PR:

$ ./check_es_system.sh -H myes.eu-central-1.aws.cloud.es.io -u monitoring -p secret -S -P 9243 -t disk
ES SYSTEM OK - Disk usage is at 9% (41 G from 450 G)|es_disk=44708304152B;386547056640;459024629760;0;483183820800

When applying the PR:

$ ./check_es_system.sh -H myes.eu-central-1.aws.cloud.es.io -u monitoring -p secret -S -P 9243 -t disk
expr: non-integer argument
expr: non-integer argument
expr: non-integer argument
expr: non-integer argument
./check_es_system.sh: line 390: [: 44708304152: unary operator expected
./check_es_system.sh: line 393: [: 44708304152: unary operator expected
ES SYSTEM OK - Disk usage is at % (41 G from  G)|es_disk=44708304152B;;;0;null

Maybe the filter needs to be adjusted. Can you check please?