YakShaving: Shawn Ferry's Weblog
v. intr. [MIT AI Lab, after 2000: orig. probably from a Ren & Stimpy episode.] Any seemingly pointless activity which is actually necessary to solve a problem which solves a problem which, several levels of recursion later, solves the real problem you're working on.
Archives
« October 2006 »
SunMonTueWedThuFriSat
10
14
15
18
20
21
22
24
25
27
28
29
30
31
    
       
Today

 Subscribe

Search

Links
 

Today's Page Hits: 319

Locations of visitors to this page
« Previous day (Oct 8, 2006) | Main | Next day (Oct 9, 2006) »
20061009 Monday October 09, 2006
Sun Provides Universal Access to Safari Books Online
So Sun has negotiated access for all Employees to the Safari eBook Library.

I have used and been happy with the Safari Books before...unfortunately I find myself slightly disapointed by the following

Unable to generate a temporary class (result=1). error CS2001: Source file 'C:\WINDOWS\TEMP\pvpviplu.0.cs' could not be found error CS2008: No inputs specified

Not that it didn't resolve itself in a moment (presumably after the file was copied into place) but the use of windows in the backend.

Shame on you Safari Books Online, of couse if I worked for Microsoft I would probably be happy.

Maybe we are working out a trade, we get access to your books and we will help you move to our wonderful Solaris operating system.
In any case, Sun Managed Operations would be happy to help you implement the migration AND monitor the entire infrastructure BOTH running on windows and running on Solaris.

Go and Register Yourself for access: Sunlibrary
I must say, that was one of the easiest self provisioning processes I have experienced in a very long time.

Technorati Tags: ,


Oct 09 2006, 01:51:04 PM EST Permalink Comments [1]

CEC: In Review
CEC2006: Too much content not enough time

Next Year:

Interactivity(Web2.0):
Suggestions:

CEC Posts in Review:
I was going to collect links to categories and such or individual bloggers/entries...too much work so instead
Technorati: cec2006



We have an internal survey on this year's CEC, I should go do that. If there is one thing I am generally not lacking, it is opinions.

Technorati Tags:


Oct 09 2006, 10:00:08 AM EST Permalink

CEC: Managing Systems at Grid Scale (Our Presentations)
Shawn Ferry (ME) and Brian Smith


In general our  attendance was a little bit lower than I had been hoping for. I think we got in the neighborhood of 80 total...I think I should be able to find out exactly somewhere.

Both instances went fairly well we had the required Managed Operations content. In the future we need to go straight to an elevator pitch and be done with it. I watched as people got bored hearing about how managed operations does things. Certainly they were interested in the more technical details CTA, encrypted transport and such. Unfortunately in our general Managed Operations presentation the less technically interesting details are more prevalent and this is a geek conference.

Both presentations had a few Managed Operations staff in attendance for solidarity. As well as a couple of new Managed Operations staff from APAC and EMEA.

The first presentation had a slight timing problem, the 15 min max planned managed operations specific content ran about 30min, in the end I got my content in with about a minute to spare and Brian got the last bits in about 5 after the scheduled end.  

The second presentation went more to plan. I was able to go into more depth on the technical aspects of the monitoring. For those who attended the presentation. We send an alert.

The only real failure for the whole thing was the attempt for a demo in the extra 5min before the scheduled end of the second presentation during the question and answer period. I had maintained what looked like good network connectivity until I tried to use it, at which point it all went pear shaped and I lost network access again.

In brief we covered monitoring methodology:
We have a top down and bottom up approach.

Top Down(A Holistic View):
We start measuring from a user experience, can you reach the web site.
Then a complex web site walking script/library(That I wrote) performs a:

At various points in the process if we don't get specific results (Welcome to the Grid,Job X submitted, Job X Successful, Logout Complete) or the process is taking too long we send an alert.

Really in the Middle Here(Business Process Monitoring):
The from a Managed Operations Point of view what we want to know is what went wrong.

A user can get an login failure for a number of reasons:
A job can fail for a number of reasons:
Some of these things we can see cause from the middle and some we need a lower level view.
Bottom Up (A small sampling, we monitor a very wide range):

Examples:

Presentation Things I learned:

Technorati Tags: , , , , ,


Oct 09 2006, 09:20:30 AM EST Permalink

Blog Information Profile for YakShaving