NCPA agent issue

[ssax]

That 500 error is likely coming from the remote NCPA agent.

Please enable debugging in your ncpa.cfg then restart the ncpa_listener service.
- Make sure you set the one under the [listener] section

Code: Select all

loglevel = debug

Then after restarting the ncpa_listener service, run the same test command to reproduce the 500 error a couple times:
- Make sure you change X.X.X.X and YOURTOKEN

Code: Select all

python check_ncpa.py -H X.X.X.X -t 'YOURTOKEN' -P 5693 -M cpu/percent -w 20 -c 40 -q 'aggregate=avg'

Then attach your /usr/local/ncpa/var/log/ncpa_listener.log.

Thank you!

[krista611]

We are experiencing the same issue on Solaris 12.

On a Solaris 12 host, we have installed the agent package (2.2.1) and are able to validate the agent listener function via browser and login. However, Nagios is unable to connect. Test attempts produce a server error 500 response. We are able to telnet, from Nagios host, to the host agent ip/port and confirm there is not firewall in place.

I’ve examined the listener log, attached, and found lines corresponding with Nagios test connects indicating an attribute issue?

File "/usr/local/lib/python2.7/site-packages/werkzeug/local.py", line 347, in __getattr__
AttributeError: 'Request' object has no attribute 'is_xhr'
2020-05-27 15:00:42,728 10014 INFO ::ffff:172.12.12.191 - - [2020-05-27 15:00:42] "GET /api/interface/ HTTP/1.1" 500 3980 0.032672
2020-05-27 15:00:42,772 10014 ERROR Exception on /api/disk/logical [GET]

[cdienger]

Hi @krista611,

We have a ticket open with you for this issue and will continue working through there.

@keerthi.seetharaman, please provide the logs that ssax has requested when you can.

[ssax]

We don't currently build for Solaris 12 but please try re-downloading and re-installing the package, the developer said he rebuilt both the sparc and i86 versions of the package just yesterday to deal with that error and you should give it a try.