Leadership By Numbers :: Steps for Linux Sametime 8 Clustered Failover (with PDF) !

Knowledge is not power, but learning is. As kids, we made art with paint-by-number kits. Simply matching the outline numbers with an oil paint gave us the illusion of mastery. I've worked in all facets of the IT industry: as IT management, as a consultant, systems adminstrator, developer and instructor. And, I still see a lot of paint-by-number projects which try to be the real thing. IT leadership is about reading the numbers, then going outside the lines and taking risks.

LEADERSHIP
BY NUMBERS

IBM/Lotus/Linux/Java
for the DC Metro region

HIGHLIGHTS

OF THE ADMINISTRATION GUILD

SECURE YOUR DOMINO SERVER ID

PDF OF LOTUS KEYBOARD SHORTCUTS

CREATE AN EASY SECURE PASSWORD

Domino Security
Eclipse
Linux
Management
Open Source
Programming
WebSphere

PDF OF BP305:Retro To Rocket

BLOGGED

« Global Accounts that use Lotus Notes with Wordle ! | Main| Ouch. Sun Microsystems Might Become a Disruptive Force »

Steps for Linux Sametime 8 Clustered Failover (with PDF) !

Category Administration Sametime

Creating A Sametime 8 Clustered IM Architecture

Here is my how-to for creating Sametime IM fail-over architecture. I'm running my Sametime 8.01 cluster on RedHat 4.4 servers with Lotus Domino 7.03. I've Been very pleased with the response time and low use of system resources.

However, it's tough to configure a Sametime 8 cluster when relying on round-robin DNS (rather than an IP Sprayer) to handle the TCP/IP connection redirects. This documentation should clarify all the details that tend to go missing in the official documentation.

You're going to save weeks of troubleshooting.

Technorati Tags: Lotus Sametime, Instant Messaging, IBM/Lotus Domino, Linux

Posted by Jack Dausman On 07/01/2008 | Permalink

Comments

2 - Thanks, Tim. What I didn't really say in the article, was how many permutations of our setup I built. A lot. The Linux implementation of ST was rough, and it wasn't until 8.01 that all the details finally clicked (e.g., Mac client download, ST web interface, etc.).

I did try the all-in-one naming standard, as you suggest. In our infrastructure, it never worked. It might be that when we untangled the difference of the hierarchical and flat usage, we had solved the real problem and the all-in-one naming would have worked.

The servers are now rated as "production" so I can't reconfigure them without paperwork, but I'll keep your suggestion for the next occasion.

Posted by Jack Dausman At 04:41:35 AM On 07/01/2008 | - Website - |

1 - Good article. One tip, I'd like to add is that for the location document problem. It's easiest to set everything to have the same name, then there will be no worries with how DCC pulls down the information, rather it come from the person document or the policy document.

For example,in stconfig the cluster document could be:
Cluster Name: server.acme.com
DNS Name: server.acme.com (of course this is the DNS entry name or ip sprayer)

and the same in the person document for the sametimeserver field:
sametime.acme.com

HTH,
Tim E. Brown

Posted by Tim E. Brown At 02:25:11 AM On 07/01/2008 | - Website - |