DX Infrastructure Manager

  • 1.  Best Practices for multiple nas

    Posted 12-03-2015 12:35 PM

    Hello, currently I have a multi hub environment with a single nas on the primary hub. I am looking to set up nas on the remote hubs to handle alarm manipulation and filtering before it gets back to the primary hub. This is to reduce the load on the processing of alarms at the primary hub.  I am just looking for any best practices when setting up this way, recommendations or things to watch out for. I would appreciate any feedback anyone may have.

     

    Thank you



  • 2.  Re: Best Practices for multiple nas

    Posted 12-03-2015 03:06 PM

    I only use two nasses, one on primary and one on backup. With that background, I don't have much to suggest in terms of best practices. I have, however, one some occasions noticed that the two nasses can get out of sync if you replicate alarms between them. I've also heard some people complain about having bigger issues with this, so that's something I'd definitely keep an eye. Maybe create a script to monitor differences between nasses.

     

    -jon



  • 3.  Re: Best Practices for multiple nas

    Posted 12-06-2015 01:41 PM

    I run four nas probes - one pair on Linux HA hubs that receives all alerts and applies preprocessing . Then a replication rule moves that to a second pair of nas probes on my Windows HA pair central/primary hub.

     

    I have it set up such that the two HA pairs have bi directional replication.

     

    Linux to Windows is uni-directional plus acknowledgements back.

     

    All four nas probes are set up with the same preprocessing scripts. AO execution is managed by the HA probe.

     

    It works though I have been told by support that this is not a supported configuration. In my interpretation that means that it has not been tested but there's no documentation to indicate that it shouldn't work.

     

    Getting things in sync is a chore and if they ever get out of sync you essentially have to manually correct the differences or delete and recreate.

     

    The biggest issue I've run into so far (when the nas is actually doing what it is supposed to) is that there is no concept of replication conflict resolution. So, if you have selected on one nas a handful of alerts and acknowledged them, then on the other nas an update occurs, these messages have the chance of passing on the network because replication sends batches with a polling delay. So you could acknowledge an alert on one nas, see it disappear, then a short time later it is back because of the update on the other nas.

     

    This is probably working as designed but it creates a serious hole in functionality. As such you really want to make sure that you design where your update occur such that you minimize the number of places where updates could conflict. eg, if you ingest alarms on one nas and run a "close if created more than 24 hours ago" script on the other, then you will have problems.

     

    -Garin



  • 4.  Re: Best Practices for multiple nas

    Posted 02-08-2018 07:14 AM

    Hello Hazard!

     

    Do you still have the same question?

     

    If you need more info about Secondary NAS, here we have some Documents that may help you:

     

    https://support.ca.com/us/knowledge-base-articles.tec000003173.html 
    https://support.ca.com/us/knowledge-base-articles.TEC000005070.html 
    https://support.ca.com/us/knowledge-base-articles.tec1967715.html 
    https://support.ca.com/us/knowledge-base-articles.TEC1816578.html 
    https://support.ca.com/us/knowledge-base-articles.TEC1580452.html 

     

    Please mark the correct answer if you already have the answer, or let us know if you need any more informations.

     

    Kind Regards,

    Alex Yasuda
    CA Technologies
    Sr Support Engineer