4 Replies Latest reply on May 22, 2018 3:09 PM by VijayaBhaskar

    Scan off tags with attribute "Archiving=1" & its impact on the Archive Size

    VijayaBhaskar

      Hi All,

      We have a situation where more than 150 K Tags are scanned off (SCAN=0) due to several reasons and while analyzing the tags configuration we observed that for Archiving set to 1 for most of these tags.

      Here we would like to understand impact of the archive size due to these tags , as per my understanding if we set archiving=0  for these tags then during archive shift PI server will not create a Primary Record and will have more space available for secondary records for the other tags.

       

      Please let me know your comments and appreciate your help.

       

      Regards

      Vijay

        • Re: Scan off tags with attribute "Archiving=1" & its impact on the Archive Size
          afink

          Hi Vijaya,

          Even if the archiving attribute is set to "0" for a given tag, a primary record will still be initialized in the archive both during point creation, or during an archive shift.  Generally, a primary record is always initialized for every point created on the server regardless of the tag configuration.  However, obviously when the archiving attribute is set to "0" == "OFF", then no incoming data will be archived while this is the case, however, the primary record is still initialized so that if archiving is re-enabled, the archives will be ready to store data for the point.

           

          Why are you concerned about the archive size with regards to these points?  Is there valid historical data for these tags which are all set to scan off?  If so, does this data still need to be accessible to users?

           

          - Adam

          1 of 1 people found this helpful
            • Re: Scan off tags with attribute "Archiving=1" & its impact on the Archive Size
              VijayaBhaskar

              Hi Adam,

              Thanks for your response , now I understand.

              Here my concern is with the limited disk space available with us and cost for procuring new storage space.

              As I mentioned , we have around 150K tags and if PI server creates a Primary Record for each tag then these tags occupy around ~146 MB on each archive, this a concern for us.

               

              Regarding the historical data and usability , that is something we haven't started analyzing yet.

               

              Nishanth Devarakonda

              Regards

              Vijay

                • Re: Scan off tags with attribute "Archiving=1" & its impact on the Archive Size
                  afink

                  Hi Vijaya,

                   

                  No problem at all, and to follow up, what is your current archive size?  You may want to consider increasing this size.  Because the primary region of ~146 MB is fixed for each archive, the less archives you have (the larger the archives are) on the server the smaller the footprint this primary region will have on disk space.

                   

                  I'm also somewhat surprised that cost for adding more storage space is a potential issue, as even multiple TB hard drives can be found for quite cheap.  Unless your servers are hosted by a third party virtualization platform, then perhaps there are additional costs associated with this since I am not as familiar with these offerings.

                   

                  - Adam

                    • Re: Scan off tags with attribute "Archiving=1" & its impact on the Archive Size
                      VijayaBhaskar

                      Hi Adam,

                      Apologies for the delay in responding to you message.

                      The Archive size is 4 GB but we have a problem with 146 MB of archive is not being used  , on further analysis we found that several of these tags were created by by mistake / bad configuration , currently we are working with stakeholders to delete these tags.

                      In regards to storage , yes I agree with you on the hard disk pricing but even I do not have knowledge on the exact virtualization platform or the costing so will not be able comment further.

                       

                      Currently we are in the process to clean the PI Data archive by fixing the tags , security , will let get back if we need more information or help.

                       

                      Thanks & Regards

                      Vijay