Right.

You do win when you have multiple clusters coming into a gmetad (we can choose one cluster out of many), but when the clusters get really big, you're right, we want a way to find the hostnames without getting their metrics. We could add a different kind of filter, such as:

/Onyx?filter=directory

or something like that.

On Thursday, May 1, 2003, at 10:17 AM, steven wagner wrote:

The problem is that this requires the web front-end to already have a host list for the cluster (which it's going to get from the cluster slice view, which has metrics in it... back to square one).

I suppose it would be possible to implement an alternate XML parser for just building the host/cluster/grid lists...


Federico Sacerdoti wrote:

It can do a host view of this type: (request: /Roscoe's Chicken and
Waffles/fryer01)

<CLUSTER NAME="Roscoe's Chicken and Waffles">
<HOST NAME="fryer01" REPORTED="sometime recently" IP="127.0.0.1">
 <METRIC />
 ...
</HOST>
</CLUSTER>

So it can give a single host-only view, and I do hope it will helf speed
things up on the web site.

Federico

-----Original Message-----
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] Behalf Of
steven wagner
Sent: Wednesday, April 30, 2003 5:46 PM
To: Federico Sacerdoti
Cc: Ganglia
Subject: Re: [Ganglia-developers] Announce: subtree-capable gmetad


Oooooooh!

I haven't played with this yet (I'm examining the diff right now) ...
but I really hope a host-only cluster view is in there.

In other words:

<CLUSTER NAME="Roscoe's Chicken and Waffles">
   <HOST NAME="fryer01" REPORTED="sometime recently" IP="127.0.0.1">
   <HOST NAME="fryer02" REPORTED="sometime recently" IP="127.0.0.2">
   <HOST NAME="fryer03" REPORTED="an hour ago" IP="127.0.0.3">
</CLUSTER>

The summary code looks like it summarizes all numeric metrics for the
cluster but doesn't display the individual hosts...

I can tell you something like the above would dramatically speed up the web front end's response time in my situation - don't know about others. :)

I may be too busy to actually run this right now, but I'm not too busy
to be grateful.  Thanks Fed! :)

Federico Sacerdoti wrote:


I have just committed some work I have been doing recently on gmetad.
The version currently in cvs supports interactive requests on port
8652. I plan to adapt the webfrontend to take advantage of this new
mode for much higher performance soon.

_______________________________________________
Ganglia-developers mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ganglia-developers



-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
_______________________________________________
Ganglia-developers mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ganglia-developers





-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
_______________________________________________
Ganglia-developers mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ganglia-developers

Federico

Rocks Cluster Group, SDSC, San Diego, CA


Reply via email to