Comments on: Sneak Peek: Scorpio Server Monitoring

By: Ashwin

Ashwin — Mon, 05 Mar 2007 06:03:49 +0000

Thanks for calling that out, Sana - I had some problems with my blog over the weekend, had to rebuild it, and missed the presentation. It’s up now.

By: Sana

Sana — Sun, 04 Mar 2007 17:22:41 +0000

Hi,

I think the link does not work,
http://blogs.sanmathi.org/ashwin/downloads/max2006preso/

Have a look, If this works then let me know.

Thanks

By: Umer Farooq

Umer Farooq — Wed, 06 Dec 2006 18:16:30 +0000

Sweet! can’t wait for this… Any chance of seeing some of the image manipulation functionality.

By: ashwin

ashwin — Sat, 18 Nov 2006 11:56:22 +0000

Thanks, Diarmuid!

By: Diarmuid King

Diarmuid King — Fri, 17 Nov 2006 20:37:23 +0000

Hi Ashwin,

Great presentation. I cant wait to get my hands on the monitor. It will be amazingly useful to get such a clear window into the running processes in production.

Thanks,

Diarmuid

By: ashwin

ashwin — Thu, 09 Nov 2006 06:30:25 +0000

@Dan, my pleasure.

@Brandon, @Mike, all good ideas that we have talked about. I’ll pass on your feedback.

By: Mike Kear

Mike Kear — Wed, 08 Nov 2006 17:55:04 +0000

What provisions are there for site owners to get a look at some of the information about their sties in a shared environment?

Obviously, if you own one or more sites on a shared box, you shouldnt expect to get rights to see the monitoring of the whole server, but some of those templates would be very useful at a site level - the query performance stuff, the request performance stuff.

Is there any provision for shared hosting performance monitoring at the site level?

By: Brandon Harper

Brandon Harper — Wed, 08 Nov 2006 17:48:57 +0000

Thanks for posting, Ashwin– definitely good information.

A couple of things I noticed–

1. There doesn’t seem to be a feature to kill hung threads automagically which in my experience are the biggest reason that servers become unresponsive (at least in the applications I work with). Basically the hung threads start stacking up over time until there are no threads available to process requests– in particular this seems to happen with flaky database drivers and other 3rd party service calls which do not return results. Have you guys thought about creating something that will kill hung threads over x seconds old, etc?

2. This looks great on a per server level, but is there anything that allows you to monitor/administer many servers in an enterprise from one spot, or is this just on an individual server basis?

By: Dan G. Switzer, II

Dan G. Switzer, II — Wed, 08 Nov 2006 14:43:57 +0000

Thanks for posting this. It’s very helpful for those of us who were unable to attend MAX this year.