Forum Discussion

lubosst's avatar
lubosst
Frequent Visitor
5 years ago
Solved

PBI On Premise Data Gateway Errors (minimizing using system resources)

Hi, I was playing with gateway settings to help our server, which is not the fastest one. I try to modify MemoryUtilizationPercentageThreshold and CPUUtilizationPercentageThreshold and set them to 60%.

Then, when my dataset in PBI services was trying to refresh, I got error: DM_GWPipeline_Gateway_GatewayOverloaded. It was easy to figure out, where is the problem and retun default values in config file.

 

But interesting thing, I was unable to find anything about this error. Even google did not find anything ;). Is there some list of all possible errors Gateway can have?

What recommendations do you have to minimize system resources for Gateway?

 

(Memory and CPU limitations lead to errors, Mashup Containers limitations results in report didn't even start to refresh - in services, gatway looks like offline)

 

Thanks for help.

  • Anonymous's avatar
    Anonymous
    5 years ago

    Hilubosst

     

    Gateway sizing recommendations depend on many variables. like Cache data workloads,Live Connection and DirectQuery workloads.

    We provide you with general recommendations that you can take into consideration.

    For more information,you can refer to this:On-premises data gateway sizing  

     

    On the other hand,enable addtional logging setting  will increase the log size significantly, based on gateway usage. We recommend that after you finish reviewing the logs that you disable additional logging. We don't recommend leaving this setting enabled during normal gateway usage.

     

    Hope it helps.

     

    Best Regards,

    Caitlyn Yan

     

    If this post helps then please consider Accept it as the solution to help the other members find it more quickly.

     

     

2 Replies

  • "What recommendations do you have to minimize system resources for Gateway?"

     

    You're holding it wrong.  You want "moar power", not less.  Use a gateway cluster with multiple cluster members. Each cluster member should have at least 4 cores (better 8), at least 8GB RAM (better 32) and ample of fast disk space.

  • Anonymous's avatar
    Anonymous
    Not applicable

    Hilubosst

     

    Gateway sizing recommendations depend on many variables. like Cache data workloads,Live Connection and DirectQuery workloads.

    We provide you with general recommendations that you can take into consideration.

    For more information,you can refer to this:On-premises data gateway sizing  

     

    On the other hand,enable addtional logging setting  will increase the log size significantly, based on gateway usage. We recommend that after you finish reviewing the logs that you disable additional logging. We don't recommend leaving this setting enabled during normal gateway usage.

     

    Hope it helps.

     

    Best Regards,

    Caitlyn Yan

     

    If this post helps then please consider Accept it as the solution to help the other members find it more quickly.