Project and Portfolio Management Practitioners Forum
cancel
Showing results for 
Search instead for 
Did you mean: 

A rule throw error when it goes to cluster

Highlighted
sevcan
Collector

A rule throw error when it goes to cluster

Hey all,

 

We have a cluster system. When the request goes to the cluster a rule gave error. But when we open it with the main server, it works fine.

The error in the below

 

   

Wrong number of values returned by SQL

in Rule 111 - {1}:

select '1','1' from dual where exists(select status_code from kcrt_requests where request_id in

(select parameter1 from knta_references

where source_id='101777'

and target_type_code='20')

and request_type_id=(select request_type_id from kcrt_request_types where request_type_name='Technology Impact Approval')

and status_code='IN_PROGRESS') (KNTA-10519)

GUID=D50C4EE8-528C-08E0-A654-C348994389C9

GUID=D67B29F7-D402-5D67-F2B6-6809CA1EB7EA

 

What can be cause of this?

Sev

6 REPLIES
Mohit_Agrawal
Frequent Visitor

Re: A rule throw error when it goes to cluster

Hi..

 

I just noticed one thing that you are missing right parenthesis at the end of SQL query. Just try once with below..

 

=================================

 

SELECT '1', '1'
  FROM DUAL
 WHERE EXISTS (
          SELECT status_code
            FROM kcrt_requests
           WHERE request_id IN (
                        SELECT parameter1
                          FROM knta_references
                         WHERE source_id = '101777'
                               AND target_type_code = '20')
             AND request_type_id =
                     (SELECT request_type_id
                        FROM kcrt_request_types
                       WHERE request_type_name = 'Technology Impact Approval' ) )

 

 

=================================

 

 

Thanks!!
Mohit Agrawal

::Please mark KUDOS and ACCEPT as solution, if this resolves your query..

pepdwill
Senior Member

Re: A rule throw error when it goes to cluster

I also encountered this exact same issue yesterday in our test environment cluster.  Very strange - the rule works on the main server but I get the same "Wrong number of values" error on the clustered server.

 

I am thinking this is symptomatic of a different issue though.   Before I got the error - I noticed other strange behavior happening.  For instance, I added a field to a data source definition in workbench, and when I went to configure the portlet definition to include this field from the cluster server, the new field wasn't there.  But it was there on the main server.

 

 

Danny

 

 

pepdwill
Senior Member

Re: A rule throw error when it goes to cluster

FYI -  bouncing the server seemed to fix the error for me.

Jim Esler
Honored Contributor

Re: A rule throw error when it goes to cluster

What PPM level are you running? We saw problems when running more than one node on a server with 7.5. The nodes lost the ability to communicate via multicast after about a week of operation. Multicast messages are used by the nodes to indicate changes that affect cached entities. Loss of multicast communication means other nodes could use obsolete information from their caches. I believe the multicast communication processes were completely redone in 9.1 but we have not tried running that configuration after our upgrade.

pepdwill
Senior Member

Re: A rule throw error when it goes to cluster

Good info Jim.  I encountered the issue on 9.14.

TorstenNeumann
Super Collector

Re: A rule throw error when it goes to cluster

If you change a rule in a cluster setup, you need to clear the rules cache after saving the request type. If you don't, old erroneous versions of the rule can still exist on other cluster nodes. That's why a restart solves the problem but you could also try:

kRunCacheManager.sh 35

(35 = Rules on our HP PPM 9.32 instance - on other versions this number may be different. Try "A" if you are not sure)

//Add this to "OnDomLoad" event