3 Replies Latest reply on Aug 3, 2018 3:46 PM by jyi

    Timeout error in PI Data Archive

    Suganya_Ganesan

      Hi All,

       

      I could see number of occurrences of below error message in PI Server logs. This happens with other tags as well. My understanding is that this happens when the Archive subsystem is trying to flush the event to archive file while there is a read lock in the corresponding tag.

       

      T:7408 PT(ptid=416209,arcsetno=0,recno=28650,loc=10,flag=8): ptwait tick/timeout=5429/120000, RID|LLL|LC|LI|NAL=0|2|6|0|249 Lock[251].uc=1 QL.sc/tid=1/0 threads=8,7408,0,0,0,0,0,0,0

      I am interested to know:

      1. If this has any negative impact on the PI Data Archive or to the events that are flushed to the archive file.

      2. Is there a way i can identify the process/source which is locking these points?

      3. Is there any action that needs to be done to prevent these messages?

       

      Additional details:

       

      1. PI Data Archive Version: 3.4.405.1198 (2016 R2). It is a PI Collective and could notice the error in both the members of the collective.

      2. The data frequency for these tags is in seconds(10 sec, 30 sec,...) and the data is fetched through OPC HDA/DA Interface.

      3. Tuning Parameter values:

          

       

      Please let me know if you require any additional details to clarify this. Thanks in advance!!

       

      Regards,

      Suganya

        • Re: Timeout error in PI Data Archive
          Lal_Babu_Shaik

          Hi Suganya

           

          Please refer KB00135 - Troubleshooting PI Archive Subsystem point lock issues

           

          https://techsupport.osisoft.com/Troubleshooting/KB/KB00135

           

          Regards,

          Lal Babu

          2 of 2 people found this helpful
            • Re: Timeout error in PI Data Archive
              Suganya_Ganesan

              Hi Lal,

               

              I read this KB but couldn't find the details that I am looking for. Can you please help me in clarifying that.

               

              Regards,

              Suganya

                • Re: Timeout error in PI Data Archive
                  jyi

                  Using your message and KB00135, below is the interpretation to your message:

                  T:7408 PT(ptid=416209,arcsetno=0,recno=28650,loc=10,flag=8): ptwait tick/timeout=5429/120000, RID|LLL|LC|LI|NAL=0|2|6|0|249 Lock[251].uc=1 QL.sc/tid=1/0 threads=8,7408,0,0,0,0,0,0,0

                  The wait time for this operation is 5.429 seconds out of lock timeout(120 sec). The thread requested a point lock on the point with pointID=416209 in order to read(flag=8) the data for this point.

                   

                  So in fact, the message is printed because the write operation took in place before your read operation. Your read operation is shown as your flag=8. This is different from the example in KB00135 in that it has flag of 4, which is flush/ point edit operation.

                   

                  To answer your question:

                  1. If this has any negative impact on the PI Data Archive or to the events that are flushed to the archive file.

                  > No. The message is mostly informational and can occur during normal operation. Also, the message does not mean that each operation (read or write) was not successful. But if you continue to see the message often, then it could mean either (shortage in hardware resources or indication of rouge/heavy user)

                  2. Is there a way i can identify the process/source which is locking these points?

                  > Not with the above error message only. You have to compare with the pi archive subsystem's thread information. Please read 3223OSI8 to know more about what thread is used by what Connection ID. With connection ID, you can identify which application/user initiated the operation.

                  3. Is there any action that needs to be done to prevent these messages?

                  > Only if the message is printed very often. Please call our tech support line if you do find this concerning. The root-cause can be different depending on the overall picture of your archive server.

                   

                  Best,

                  Jin

                  2 of 2 people found this helpful