none
Virtual Array Tiered Storage Policy RRS feed

  • Question

  • I have a small proof of concept running at my agency, and while I am trying to write up some of the documentation I am trying to figure what the triggers are for moving data out to the Azure storage account?

    I have one device setup for file shares, and I am over the 10% local storage on it, I would have though that would have triggered data move but instead I just got an alarm in the Azure console about using too much local?  I would think space wise the data would be moved out of the local array when I got close to the 10% local storage number?

    I just need to know what triggers (time not used?  space considerations?) the moving of data out to cloud.  I see the backups of the local array go out no problem, so I know I am connected.



    Dave Lawlor
    I offer advice as is, and you use that advice at your own risk. I am not in your environment so I may not infer everything that is going on there from your post/question. Use common sense if you know this advice conflicts with your environment.

    Thursday, March 15, 2018 6:14 PM

Answers

  • Hi Dave,

    Can you go to the local UI and get the logs and email those to me (amason@microsoft.com)? I'll get someone to take a look and see if they can figure out what is going on.

    Thanks,

    Andrew

    Wednesday, March 21, 2018 11:31 PM

All replies

  • Hi Dave,

    The documentation for this is here: https://docs.microsoft.com/en-us/azure/storsimple/storsimple-ova-overview#automatic-storage-tiering, it is heat map based and depends on usage and age of data.

    Hope that helps,

    Andrew

    Monday, March 19, 2018 4:55 PM
  • I have seen that documentation but it doesn't give any kind any real details.

    In my current test, I have provisioned 1TB, so that is ~100GB locally for 10%. 

    On Feb 12th I placed about 2 dozen files totaling about 60GB
    Saw the backup traffic of the local share that night go out ~12GB with Dedupe

    Nobody has access to these files and there are no local backups for this device, so I know they have not been access or touched by any user.

    2 weeks ago I put another 46GB of files onto the same share, again saw the backup traffic spike that night.

    Now I have a share that is at 106GB and received an alert I was over the 10%, but nothing is getting moved to cloud yet (alert was 2 weeks ago).

    If files haven't been touched since Feb 12th over a month ago and still not tiered out when the local storage is almost at capacity, I have to think something is not right?

    I want to propose this as a solution for an archiving system that is 2TB in size right now but will double over the next few years.  How in the world can I provision for that if I don't have any idea how it will move.  What benefit (besides backup) do I get if I have to provision the full amount of data locally because once I cut over the 2TB are going to sit locally forever it seems?



    I offer advice as is, and you use that advice at your own risk.
    Dont forget to mark as Answered if you found this post helpful.




    Monday, March 19, 2018 5:28 PM
  • Hi,

    It seems like it is functioning normally. The system will only spill out a sufficient amount of data from the local tiers to bring it down to about 78% local utilization. If the backups were successful it means that all the data from the appliance has already been backed up to the cloud. The warning might momentarily appear if the ingestion is happening at a rate faster than we are able to spill to the cloud.

    If you still have questions or want to dig deeper into this, its probably best to get in touch with support to look at the logs and get more insight into the system.

    Thanks,

    Andrew

    Tuesday, March 20, 2018 3:35 PM
  • Thanks Andrew -

    I am with a smaller federal agency, and we are on the trial subscription to Azure.us right now with no support contract (I was told our normal EA support wouldnt cover it), when I tried to do support in the portal it sent me to twitter Azure support, which sent me to here.  I have reached out to our contracting team that we do our normal EA through but have not gotten a response, its like I cant get MS to take our money for cloud lol.  The only way I get funding for this project is to have this POC working, and I cant get the POC working without paid support it seems.

    For weeks the local 10% of the 1TB share has still been at 106GB which seems much higher than 78% utilization.  Ill try through our EA team again

    I do appreciate you responding though, have a great day. -Dave




    I offer advice as is, and you use that advice at your own risk.
    Dont forget to mark as Answered if you found this post helpful.

    Tuesday, March 20, 2018 4:12 PM
  • Hi Dave,

    Can you go to the local UI and get the logs and email those to me (amason@microsoft.com)? I'll get someone to take a look and see if they can figure out what is going on.

    Thanks,

    Andrew

    Wednesday, March 21, 2018 11:31 PM
  • Andrew -

    Thanks so much Ill be sending them out today. -Dave



    I offer advice as is, and you use that advice at your own risk.
    Dont forget to mark as Answered if you found this post helpful.

    Thursday, March 22, 2018 11:11 AM