[Xymon] XymonPSClient CPU graphs missing over HTTPS

[Kris Springer]

The clientlog looks to be making it into the server completely.  It 
appears mostly identical except the https version contains the url and 
user/pass info at the bottom.

Could this issue be with the xymoncgimsg.cgi not processing the data 
correctly somehow?

Kris Springer


On 1/25/24 15:54, J.C. Cleaver wrote:
> Are the outputs of the clientlog otherwise identical for the system
> between the http and https submissions?
>
> Off the top of my head, the only thing I think would cause this on the
> server side would be a permanently/consistently truncated client report,
> causing the load not be scanned for hostsvc.test status generation.
>
> -jc
>
>
> On Thu, January 25, 2024 14:50, Kris Springer wrote:
>> Additional info: I just realized the 'files' test that I have defined to
>> check a few Windows folders also is not working either when using the
>> https method.  The defined Windows folders are read and appear in the
>> 'clientlog', but the Server's webpage says 'Could not determine size of
>> directory', probably because there are no 'filesizes' rrd's being
>> created.  When the client sends over http the rrd's are created and the
>> folders graph just fine on the server.
>>
>> Kris Springer
>>
>>
>> On 1/25/24 15:19, Kris Springer wrote:
>>> I'm using the XymonPSClient 2.42 on a Windows server, and all works
>>> fine except when I choose to send the data from the client to the
>>> server over https the CPU graph doesn't display on the server side. If
>>> I change the XymonPSClient to send over http the CPU graph displays
>>> correctly.  I've tried this on different hosts and results are always
>>> the same when using https.  Server is v4.3.30, but it was happening on
>>> an older version too.  Can anyone give me a clue how to resolve this?Â
>>> There's no errors happening that I can see anywhere. I looked in the
>>> rrd folder and there is no 'cpu' or 'la' rrd file being created for
>>> the hosts that use the XymonPSClient over https.
>>>
>>> Thanks,
>>> Kris Springer
>>>
>> _______________________________________________
>> Xymon mailing list
>> Xymon at xymon.com
>> http://lists.xymon.com/mailman/listinfo/xymon
>>
>