Anybody automate purge caches for their threads?

Clovertech Forums Read Only Archives Cloverleaf Cloverleaf Anybody automate purge caches for their threads?

  • Creator
    Topic
  • #49581
    Nate Kruse
    Participant

      Background – We want our clinical users to be able to update the .csv files that we use for lookup Tables (less work for me..yeah!).

    Viewing 4 reply threads
    • Author
      Replies
      • #62568
        Charlie Bursell
        Participant

          You can do what you propose but it may not always work.  What if the process is extremly busy or blocked?  The purgex may not run.

          I do this all the time.  Cloverleaf Tables are meant to be static and are not designed for dynamic tables.

          I have them put the CSV file in a known directory with a know name.  When my Tcl proc starts, I read the table into a keyed list or array (I pefer array) – be wary of duplicate keys.  I also store the last modified time (mtime).  With each invocation of the proc, I stat the file for mtime.  If different than the one I saved, I reload the file else I use the table as is.

          There is additional overhaed involved but not as much as building a new Cloverleaf table and running purgex. It makes the whole thing completely automatic.

        • #62569
          Jim Kosloskey
          Participant

            Nate,

            I think you just list all of the threads you want purged comma seperated followed by the purgex in the -c switch of the hcicmd.

            Something like this I think:

            hcicmd -p -c “thread_1,thread-2,etc… purgex”.

            However, just be aware purge caches refreshes more than just the Tables you just modified/created – and you may not always like the results.

            Jim Kosloskey

            email: jim.kosloskey@jim-kosloskey.com 29+ years Cloverleaf, 59 years IT - old fart.

          • #62570
            Nate Kruse
            Participant

              Thanks for the suggestions and the words of caution.  At least this is possible, but might not be the way we want to go now.  I’ll have to see what the team says.

              Thanks again!

              Nate.

            • #62571
              Gary Atkinson
              Participant

                Were on 5.5 rev.1  When we issue a purge caches command from Network Monitor on the process called “charges”, all threads were caches in the site and not just in the charges process where the command was issued.  So, I tried it from the command line using the syntax:

                hcicmd -p charges

                hcicmd> charges purgex

                Response:

                XLT-related caches purged

                Log says:

                [cmd :cmd :INFO/0:  charges_cmd:10/25/2007 13:46:41] Receiving a command

                [cmd :cmd :INFO/0:  charges_cmd:10/25/2007 13:46:41] Receiving a command

                [cmd :cmd :INFO/0:  charges_cmd:10/25/2007 13:46:41] Received command: ‘charges purgex’

                [cmd :cmd :INFO/0:  charges_cmd:10/25/2007 13:46:41] Doing ‘purgex’ command with args ‘

                So, the process “charges” only gets purged, correct  ❓ I need to make some changes in a table that the engine is currently using…

              • #62572
                Mark Thompson
                Participant

                  Gary,

                  You are correct.  In 5.5 Rev 1 only the GUI purge caches is broken (because it puts too many threads in the command).  The command line version works correctly.

                  - Mark Thompson
                  HealthPartners

              Viewing 4 reply threads
              • The forum ‘Cloverleaf’ is closed to new topics and replies.