none
TraceCollector: Trying to trace a ETL files travel from the Provider to Consumer RRS feed

  • Question

  • I shall write out my understanding please correct me if I am wrong and clarify couple of my doubts. Thanks in advance. Here the context is Azure Stack ASDK install. The Objective of this question is to understand how an Error Generated by the provider reaches the Azure Stack Portal the consumer:

    Consider Physical Memory as the event Provider. In ASDK I create a huge VM and as I watch the physical memory consumption reaches 99% and I see in the Dashboard Critical Error = 1.
    What has happened is the Provider has indicated the Critical Condition by writing out an ETL file into its component place.
    In the background Tracecollector the Controller which is constantly scanning all the component space gets the ETL file and dumps it in the common share. Under C:\TraceCollector\Configuration there are two ETL files.

    [QUESTION 1]: Please tell me which ETL file is applicable to Physical Memory
    "C:\ClusterStorage\Volume1\SHARES\SU1_INFRASTRUCTURE_2\Diagnostics\HealthAndMonitoring\BareMetal\AzureStack.Fabric.Health.BareMetal.etl"
    OR
    "C:\ClusterStorage\Volume1\SHARES\SU1_INFRASTRUCTURE_2\Diagnostics\BareMetal\VMSIF\VMSIF.etl"
    I will assume it is the first one, AzureStack.Fabric.Health.BareMetal.etl

    [Question 2]. In the below image you will see Buffers and Log files. Buffers are the component space and Log files is the common share point C:\ClusterStorage\Volume1\Shares\ is that right?

    

    [Question 3]: These ETL files keep growing and when it reaches its limit gets saved away. Does it get saved as BLG files? If so where will they get saved? My guess is the files in C:\ClusterStorage\Volume1\Shares\SU1_Infrastructure_2\Diagnostics\HealthAndMonitoring\BareMetalgets saved away twice a say at C:\ClusterStorage\Volume1\Shares\SU1_Infrastructure_2\Diagnostics\BareMetal\PerfCounters as BLG files.

    [Question 4]: Under C:\Windows\tracing\SDNDiagnostics\Logs every 5 minutes a 3MB ETL file is saved and a new files gets created. However SDN is to do with Self Defined Networking so I assume that is irrelevant to Physical Memory and Baremetal cases.

    [Question 5]: Last one, ETL is live while BLG is just static form of it. ETL can grow but BLG are fixed sizes. ETL is for realtime access BLG is for future. Are they all correct assumptions? Where does all the Baremetal ETL files get saved away? Assume my Baremetal file name is XXXX_AzureStack.Fabric.Health.BareMetal.2019-01-30.1.etl I was expecting XXXX_AzureStack.Fabric.Health.BareMetal.2019-01-30.1.blg but nothing came up. Kindly clarify
    Sorry for the series of question, I would really appreciate answers. Thanks a lot

    These questions are based off the article https://docs.microsoft.com/en-us/azure/azure-stack/azure-stack-diagnostics and ties back to the following image too. This also has relevance to this following presentation https://www.youtube.com/watch?v=BSwy0_pl-D4 in which there is a slide on TraceCollector:

    



    Thursday, January 31, 2019 4:53 PM

Answers

  • What exactly are you trying to accomplish by tracing the ETL files? Unfortunately I cannot say many things publicly, but I would be happy to follow up with you via Email. 

    Can you send me an Email at AzCommunity@Microsoft.com with a link to this post, and we will talk further?

    Thursday, February 7, 2019 10:19 PM
    Moderator

All replies

  • I am following up with this question, and will get back to you. 
    Thursday, January 31, 2019 10:15 PM
    Moderator
  • @TravisCragg_MSFT any luck? Please update thanks
    Wednesday, February 6, 2019 4:12 AM
  • What exactly are you trying to accomplish by tracing the ETL files? Unfortunately I cannot say many things publicly, but I would be happy to follow up with you via Email. 

    Can you send me an Email at AzCommunity@Microsoft.com with a link to this post, and we will talk further?

    Thursday, February 7, 2019 10:19 PM
    Moderator