I am using Jdbc adapter Notification (update notification ) in a Clustered environment .I was successful in using notification in Devlopemt box (which is not In cluster) is working fine.
But when I use Update Notification in Cluster its not Working for me .
Server A and server B are in cluster and when I am able to enable that Notification though Admin screen for Server A
And when I enable the Notification in Server B its giving me this Error :-
[ART.116.3038] Adapter Runtime (Notification): Error in Notification Callback:enableCallback notification HarmonWIN.GP.Notification:EAIControNotif.
[ADA.1.316] Cannot execute the SQL statement “CREATE TABLE WMBEAI_N1_ba333jn (xxxx int, xxxxx varchar(50), xxxx varchar(50), xxxx varchar(256), xxxx int, xxxx decimal(15,2), xxxx datetime NULL, xxxx varchar(18), xxxx datetime NULL, xxxx varchar(40), xxxxx varchar(40), xxxx varchar(40), xxxx int, xxxx NULL, xxxx numeric(9) xxxx, xxxx numeric(14))”. "
(42S01/2714) [wm-cjdbc33-0009][SQLServer JDBC Driver][SQLServer]There is already an object named ‘WMB_N1_ba333jn’ in the database."
[wm-cjdbc33-0009][SQLServer JDBC Driver][SQLServer]There is already an object named ‘WMB_N1_ba333jn’ in the database.
I am assuming that its is trying to create table again when Server A had already created .
If any one had gone though similar problem.
I will appreciate to share your thoughts and experiences…
is there any FIx or work around for this problem /???
Thanks
In 6.01 and JDBC 6.0.3, polling notifications are not cluster aware. There is a version out somewhere (either planned or GA I do not know) that corrects this issue.
What we have done is enabled polling notifications equally among our clustered servers.
thanks Bill
we are using 6.1 version and jdbc 6.0.3 version.i guess that they might not have corrected that Cluster Problem yet.I am trying to talk to webMethods they might help me with this present issue .
Is is possible for you to explan me more detail how did you guys enable polling notifcation in cluster servers.
I just make sure that they are spread out over my three servers. I enable the Adapter Connection on all three servers and then I the developer has three Polling notifications in their packages, I’ll enable one on each server… or I migh group a set of transactions on a single server… just depends on the situation, but I try and have each of my three servers “hosting” close to 1/3 of the polling notifications.
You always have a problem if the server that hosts the polling notifications goes down for any period of time, but I am able to work well with the DBA’s back-up the data in the buffer tables and delete them so I can recreate them on another server. Not the best way of doing things in my opinion, but the only work around I could find.
If you find something from wM, I would certainly like to hear about it.
Kitt - What JVM and Operating Sys are you running on your IS against? Also, did you have to do anything else in addition to applying IS6.1 Fix6?
We are noticing that our notifications are randomly getting disabled on node 1 or node 2. They work fine in a non-clustered env. We have applied Fix6 and are on HP-UX B.11.11, IS 6.1, JRE 1.4.2.06