11 Replies Latest reply on Dec 9, 2016 4:16 PM by Gary Conway

    PharosUniProxy Problems

    Gary Lloyd Newbie

      Hi we are currently experiencing intermittent technical issues with our Pharos UniPrint install. Our organisation currently runs Uniprint 8.2.


      All of a sudden the event viewer seems to show The PharosUniProxy as losing network connection every 3 minutes for a second or so:


      Network connection to 'staffprint' lost

      bede - Network connection to 'staffprint' established. Offline for 00:00:14, from 10/06/2014 13:49

      Network connection to 'staffprint' lost

      bede - Network connection to 'staffprint' established. Offline for 00:00:14, from 10/06/2014 13:52


      At some point when this happens users are unable to connect via the popup client.


      Windows updated itself about a month a ago and as far as I am aware no other changes have been made.


      These issues only seem to have started to come about over the past few days.

      We have tried restarting the servers but this has not solved the issue.


      The Sql database server seems OK and the servers seem OK when pinged etc.


      Has anyone been affected by this or know what else we can try ?




        • Re: PharosUniProxy Problems
          Scott Olswold Guide



          There are a couple of ways to approach this:


          1. You may want to ensure that your Pharos Proxy Service is at the latest build: https://community.pharos.com/docs/DOC-1632#jive_content_id_Uniprint_82
          2. Also, appropriate logging of the Proxy Service (using the Pharos LogSetter tool) may provide an indication of what's going on.
          3. The Proxy Service files can become too large over time, resulting in inefficient response times and connection issues to the Pharos Database Service running on the Principal Server.


          To see if item 3 is your culprit, you can:


          1. Stop the Pharos Database Proxy Server service.
          2. Browse to C:\Program Files (x86)\Pharos\proxy (that's the default path on a 64-bit server OS).
          3. Find the three files ProxyDataCache, ProxyLicense, and ProxyTxns. Each will have a "db" extension.
          4. Move them to the desktop.
          5. Start the Pharos Database Proxy Server service.


          Since you were getting your Event Viewer messages with a high degree of frequency, check the Event Viewer occasionally. Let us know what you find out.




          1 of 1 people found this helpful
          • Re: PharosUniProxy Problems
            Gary Conway Wayfarer

            Jeff GellerScott Olswold

            Jeff or Scott:  In connection with this answered response, we are having network issues ourselves here at BPL.  My specific question in regarding the DB files.  It is mentioned that they may grow too large and start to cause issues.  How large is too large before they may start to cause an issue.  I want to look at the branches reporting the issue and look at the size of their DB files and if they are too large, try the suggestion listed here.

            Gary Conway

            Brooklyn Public Library

              • Re: PharosUniProxy Problems
                Scott Olswold Guide


                Like so many performance killers, the threshold here is highly variable. It depends upon the hardware specifications of the server, what is competing for those resources, and where in the "blob" that is a SQL table the data lives. I've seen some large Proxy database files inside a very high-functioning print server, and others that were ridiculously average causing major problems. In short, each server's pain point will have a different threshold.


                In general, if the Proxy service begins to act up and you've eliminated network connectivity as the cause, you've most likely hit that server's threshold. Like Rojaa stated earlier, the files can be managed such that you're not just waiting around for "done" to occur on an operational basis. My advice is to dial-back the Delete Age from the default on very busy servers so you never wind up in the position to delete/move files and restart services.



                  • Re: PharosUniProxy Problems
                    Gary Conway Wayfarer

                    Scott:  Thanks for such a quick response.  But I still have a couple questions:


                    1. I like that you stated once “you’ve eliminated network
                      connectivity as the cause”.  Can you provide me any assistance on that
                      what, where to check, in order to eliminate the network?
                    2. I actually looked for that registry entry and I could not find
                      it.  I even searched my entire registry hive and could not find it.
                      I figured it was because that client was on version 8.2 and we are on version
                      9.0.  It that registry still there on our system?  If so, exactly
                      where would it be found?  Or does it need to be added?


                    Thank you for your assistance.

                      • Re: PharosUniProxy Problems
                        Jeff Geller Guide

                        Hi Gary,


                        Perhaps I can add a few things to help answer our questions. You can filter the Pharos Alerts in the Pharos Administrator for the Alert # 28100. This is an Alert that gets recorded by the Proxy Service on the Signup Servers when they finally re-establish a connection to the Pharos Database Service on the Principal Server. What happens is if/when the Proxy Service fails to connect or the connection is cutoff for some reason, the Proxy Service makes a note of when this happened. Once it finally can connect again it then sends the Alert that it was offline for x minutes. if you see this Alert quite often that indicates the Proxy is having communication problems to the Principal Server where the Pharos database Service lives.


                        This is the Alert you can search for.


                        Network connection to ‘%1’ established, was offline for %2”

                        Communication to the database server named %1 was restored. The connection had been offline for the amount of time specified by %2.



                        For the Registry Key question unfortunately Rojaa's note above did not include that those keys do not exist by default. Below is a screen capture from the help file on the registry keys she mentioned.



                        -Jeff Geller

                          • Re: PharosUniProxy Problems
                            Gary Conway Wayfarer

                            Jeff:  Thank you so much.  This is an eye opener.  Every branch, every day, is having the 28100 error code. Always at the same time for each instance.  Sometimes twice a day.  Saw a few days where it happened 3 times in one day to all the print servers.  I think that qualifies as "seeing this alert quite often".  Now what do we do to correct it?  Can you assist?  What suggestions do you have?  Gary

                              • Re: PharosUniProxy Problems
                                Jeff Geller Guide

                                Hi Gary,

                                Thanks for the information. the Alert will tell you how long it is offline for. if it is a few seconds or say even a minute that is not troublesome and if the times are all around the same time for each branch then I would suspect that there is something going on that is causing it for all branches. I wouldn't think this frequency is much to be troublesome, it is more when it happens every 5-10 minutes for a more substantial amount of time each time the alert is logged.


                                How big are the files ProxyDataCache, ProxyLicense, and ProxyTxns (all with .DB file extensions) located in the directory "C:\Program Files (x86)\Pharos\proxy (that's the default path on a 64-bit server OS)".


                                -Jeff Geller