Oracle Analytics Cloud and Server

OBIEE 12C + SSAS. Timeout

Received Response
121
Views
6
Comments

Hi!

I made subject area, based by SSAS-source (Anslysis Services 2012).

In analytics, i created a analysis.

Sometimes, when i run this analysis, i get error

Error Codes: OPR4ONWY:U9IM8TAC:U9IM8TAC:U9IM8TAC:OI2DL65PState: HY000. Code: 10058. [NQODBC] [SQL_STATE: HY000] [nQSError: 10058] A general error has occurred. (HY000)State: HY000. Code: 43113. [nQSError: 43113] Message returned from OBIS. (HY000)State: HY000. Code: 46169. [nQSError: 46169] An HTTP operation to <SERVER> timed out after 60 seconds. Operation aborted. (HY000)<SQL CODE>

How fix this error?

On the server (SSAS), i set up the IIS

File "msmdpump.ini":

<ConfigurationSettings>  <ServerName>SERVER_NAME</ServerName>  <SessionTimeout>3600</SessionTimeout>  <ConnectionPoolSize>20</ConnectionPoolSize>  <MinThreadPoolSize>20</MinThreadPoolSize>  <MaxThreadPoolSize>50</MaxThreadPoolSize>  <MaxThreadsPerClient>10</MaxThreadsPerClient></ConfigurationSettings>

There timeout is 3600 sec

In repository:

Connection pool (type XMLA) haven't such setting "Timeout"

xmla_connection_pool.jpg

Where else i must change setting "Timeout"?

Answers

  • Hi Frog, Did you ever find the timeout setting in 12c? 11 doesn't have this issue.

  • Hi Mr Sausages

    No.

    I have a long correspondence with the oracle support about this theme. But we have not found any positive result

    Support-master advised me to change the setting UserPopulationQueryTimeoutSecs

    from file

    /.../ORACLE_HOME/user_projects/domains/bi/config/fmwconfig/biconfig/OBIPS/instanceconfig.xml

    in tag <Security>

    I tried change it to other values - and <<60, and >>60 - then restart OBIEE

    But it's not helped me

  • Hi ,

    From what i know the no of connections allowed should be same in both connection pool and msmdpump.ini file. You can probably try changing <ConnectionPoolSize> to 10 or Maximum Connections in Connection Pool to 20.

    Thanks,

    Rampradeep

  • Hi Frog,

    Sorry to hear you are having as much luck with Oracle Support as we are having. Very disappoint.....

    Would be handy if Oracle tested this for themselves.

    We have had to role back from 12c to 11g where we have no issues. So it is a timeout they have introduced in 12c. It is not OS related.

    Thank god Oracle do not make life support machines....

  • We tried reducing and increasing the connections, does not make any difference, quires work great on 11g.

  • Hi Mr Sausages

    > You can probably try changing <ConnectionPoolSize> to 10 or Maximum Connections in Connection Pool to 20.

    I checked many-many different cases - set different values to "Maximum connections" (in repository) and to "msmdpump.ini"-file.

    And this case too - i experimented with different values of "ConnectionPoolSize" - not helped.

    The only thing, early, after i get timeout-error, if i immediately run the report with timeout, i will get another error - 503:

    Error code: OPR4ONWY:U9IM8TAC:U9IM8TAC:OI2DL65P  State: HY000. Code: 10058. [NQODBC] [SQL_STATE: HY000] [nQSError: 10058] A general error has occurred (HY000)  State: HY000. Code: 43113. [nQSError: 43113] И Message returned from OBIS. URL: http://<SERVER_NAME>/OLAP/msmdpump.dll. The http return status is 503. The service is unavailable. (HY000)503 

    I fixed 503-error in the following way - set "MaxThreadsPerClient" (from "msmdpump.ini") value less than "Maximum connections" (in repository).

    Ideally, when "MaxThreadsPerClient (from "msmdpump.ini)" and "Maximum connections" (in repository) are equally, this should solve problems with timeout... in my case it's not working