Skip to Content

Workload Management : User Statement Prioirty

May 24, 2017 at 06:25 AM


avatar image


We're facing issues with memory and CPU consumption in HANA due to competition for resources between many sessions. I looked to Workload Management for a solution and while it has good features, I'm not sure it fits our scenario. Our high profile jobs run only for a part of the day (night time) during which they should not be disturbed. However, for the rest of the time, Reporting users should not need to be restricted by way of CPU/memory.

Could simply setting User statement priority (value: 9) to the Job DB user do the trick? Theres not much info on how this parameter works.

Any information will be appreciated

10 |10000 characters needed characters left characters exceeded
* Please Login or Register to Answer, Follow or Comment.

1 Answer

Best Answer
Lars Breddemann
May 24, 2017 at 08:56 AM

The priority is used for a relative ranking of competing resource requests. It's not as if there is a 'priority lane' in the CPU scheduler for higher priority requests.

Having said that I recommend reviewing the SAP HANA documentation and the relevant SAP notes, e.g. 2222250 - FAQ: SAP HANA Workload Management .

Other than that, be aware that using the 'user tags' in a question does not work on user names. Neither Florian nor I have been notified about this question - that's just not a feature of the current community platform.

Show 3 Share
10 |10000 characters needed characters left characters exceeded

Thanks Lars - point noted on the User tag. Its been a while since I posted a question, hence the error in usage.

On your point about relative ranking of competing resource requests, thats kind of what we need. Only when 2 requests are running concurrently, I want to specify who's the boss (in this case, the Jobs over the user reports), there are specific user IDs for each. Will setting the priority of the Job user to 9, prevent OOMs caused by regular users (priority 5) ?


The priority does not influence memory allocation but it is used for HANA internal "job" scheduling. A "job" here is not the same as the database object "job" used by e.g. flow graphs, but HANA's internal mechanism to split up and parallelize SQL statement processing.

You can find this in the Administration Guide documentation.

If you're looking into workload management, then the common recommendation to use the latest HANA release (HANA 2 SP1, currently) is even more fitting. HANA 2 has got a number of enhancements on workload management (e.g. a feature called "admission control", to prevent system overloading). It's worthwhile reading up on that.


thanks Lars