Feature_request : http probe for size of returned page

[xkilian]

We have a use case where an http API responds with data. The size of the data varies and is correlated with a business process.
There are cases where the API responds but returns no data or unexepected amounts of data. A regex based check can be used for the first case, but having the size of the page or the size of the returned data/payload would be more informative.

Request: Have the http probe return the number of bytes in the http body or page.
Related: If implementing this, it would make sense to look at the #351 : md5 request as they are closely related. (If the decision is to implement)

[brian-brazil]

We should already have the content length as a metric

…

On Sat 2 Mar 2019, 17:11 xkilian, ***@***.***> wrote: We have a use case where an http API responds with data. The size of the data varies and is correlated with a business process. There are cases where the API responds but returns no data or unexepected amounts of data. A regex based check can be used for the first case, but having the size of the page or the size of the returned data/payload would be more informative. Request: Have the http probe return the number of bytes in the http body or page. Related: If implementing this, it would make sense to look at the #351 <#351> : md5 request as they are closely related. (If the decision is to implement) — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#426>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AGyTdm11ws7m_jwalPmB9WnztbuyrIIUks5vSrCrgaJpZM4bafpI> .

[xkilian]

probe_content_length is always -1

One of my probes returns data as binary : xgoogle/protobuf
One of the probes returns data as plain HTML with some javascript

Other metrics are ok, http_version=1.1, status_code = 200, probe_success = 1, http_redirects = 0
The nagios check_http has no problem with either of these http requests and displays the content length correctly.

[BenoitKnecht]

Is this a duplicate of #337 then?

[xkilian]

Yup, just read through the original. This is a duplicate and can be closed in favor of tracking in #337 .