Sorry, was on hols for a week.
I must say I am a little confused. It seems to be DNS/hosts related though.
We have this setup:
Primary Node: server002.company.com : 10.0.0.1
DNS Aliases: stadvanced.company.com - Port 80 is default_port for advanced;
Servers: SSC (server002.company.com),
Meetings (meetings.company.com), ST Proxy (webchat.company.com), Advanced (stadvanced.company.com)
What does this mean? Have you installed the Meeting application using meetings.company.com? Did you put meetings.company.com into the deployment plan when it asked which node to install it on?
Local Hosts File entries for meetings.company.com and webchat.company.com to 10.0.0.1
You should not need to do this but I understand why you have had to do it. This should resolve to your load balanacer/WAS proxy using DNS, ideally.
Issue here is that meetings.company.com points to the WAS Proxy in DNS but is installed as the server name and host alias name for meetings on the primary. The WAS Proxy is installed as webmeetings.company.com. I’ve since created a Host Alias for webmeetings company.com which works to attend meetinsg if I only have one of the two clustered servers running !
Secondary Node: server003.company.com : 10.0.0.2
DNS Aliases: stadvanced2.company.com; meetings2.company.com; webchat2.company.com
Servers: Meetings2 (meetings2.company.com)
WAS Proxy Node: server004.company.com : 10.0.0.3
DNS Aliases: meetings.company.com; webmeetings.company.com, webchat.company.com
Servers: WAS Proxy (installed as webmeetings.company.com which can be used to attend a meeting via the Host Aliases now created.
During the install, did you install the application servers using server002.company.com, server003.company.com etc? If you used the alias name to install then you have a mess to clear up. You can install all the application servers using the actual operating system host name and then DNS controls everything else.
If this is caused by installing to the alias names then you can either a) uninstall and reinstall or b) check out
http://www.ibm.com/developerworks/lotus/documentation/sametimehostname/
I have never used the jython scripts but there is a comprehensive document that details what to do. There is a c) which involves opening a PMR and getting some guidance from IBM.
Personally, I would reinstall. You gain more knowledge and it gives you the chance to tweak your deployment and move the SSC and DB2 to it's own host and deploy WebSphere proxies on the primary and secondary nodes. It involves doubling your effort so maybe a PMR is better because you cannot get the level of support you need from the forums :(
He's what I would do using the three nodes. The WebSphere proxy will need to listen on port 80 and front all three applications. I don't know how many users you will have accessing services but you should be aware of the resources available to the node.
The following only has a single database server so you will not have any resilience at that level.
Primary Node: server002.company.com : 10.0.0.1
DNS Aliases: stadvanced1.company.com; meetings1.company.com; webchat1.company.com
Servers: Meetings, ST Proxy, Advanced, WebSphere proxy
Secondary Node: server003.company.com : 10.0.0.2
DNS Aliases: stadvanced2.company.com; meetings2.company.com; webchat2.company.com
Servers: Meetings, ST Proxy, Advanced, WebSphere proxy
SSC and DB2: server004.company.com : 10.0.0.3
DNS Aliases: none
Servers: SSC
Load balancer
DNS Aliases: stadvanced.company.com; meetings.company.com; webchat.company.com
Return to top
. . . . . . RE: Clustering v9 Meetings with sta... (~Kim Opniman 22.Sep.14)
Print this page
