Xymon Mailing List Archive search

serverside tests using data from multiple clients?

list Vernon Everett
Wed, 23 Nov 2011 08:25:41 +0800
Message-Id: <user-b5f61a4b32ff@xymon.invalid>

We had something similar at a previous client.

I am assuming that these measures are all related, and you want to see them
all in one place?
"Invent" a dummy server. Let's call it dserv1

Remember, when you send data from a client, the client name is in the data
message.
If all of these tests send their data with $MACHINE is set to dserv1, it
really doesn't matter where the data comes from, the Xymon server will
think it's data from dserv1, and treat it as such.

Now you have all the data in one place, making consolidation relatively
trivial.

Regards
     Vernon


On 23 November 2011 08:13, Elizabeth Schwartz <user-c61747246f66@xymon.invalid>wrote:
Here's our situation.

We've got some things we want to watch multiple servers, that will
come through as a name-value pair (once we write the local test)

So on server one I might have
  foo-23
  bar-25
and on server two I might have
 bar- 17
 baz 3

We want to to do some tests that would involve adding and correlating
all this information. There can be *many* possible names, not just
three

What's the most xymon-like way to do it? Can I send data through
client tests that is *read* by server-side tests ? (I vaguely remember
there was a way to do it but searching is not workign for me today).

One way to do it is to make a log file for each client, rsync the logs
to a common location , and then run a single custom test on that
location, but that strikes me as beign a bit "outside the box". I want
to work within xymon whenever possible, and having individual client
tests and the combo test accessible from the xymon interface would b e
nice.

thanks for any thoughts or clues

-- 
"While it is futile to try to eliminate risk, and questionable to try to
minimize it, it is essential that the risks taken be the right risks. "
- Peter F. Drucker