4 Replies Latest reply on Jan 26, 2016 9:04 AM by Adam Demeter

    Faulty PCs in 9.0 R2

    Adam Demeter Scout

      So, I have a strange issue going on since updating to 9.0 R2 from 8.4. We have about 170+ PCs in our library, and another 50+ PCs in a different building on our campus. When PCs (in either location) show faulty in the Nerve Center, students are not assigned a PC (which is understandable, we had this issue in 8.4). However, in 8.4 we would have maybe 1 or 2 PCs become faulty, we would reboot the PC, and all was well. Since 9.0 R2, we will have up to 30 PCs becoming faulty (random, never the same ones). We found the faulty PCs that were not being used, went into an idle state (PC never went to sleep or hibernate, and the screen was still on). Once we moved the mouse, about 1-2 seconds later, the PC is no longer faulty in the Nerve Center, and automatically assigns a student to that PC.

       

      I have been working with Pharos Tech support on the issue, but we have come up with nothing so far. I supplied them when log files from multiple PCs, and they see there is a point when the PC and the server cannot not talk to each other, and therefor the server marks the PC as faulty, but we are not sure why this is happening. Pharos says there is nothing different between 8.4 and 9.0 R2 in the way the PCs talk to the server, but I am beginning to think otherwise. I am having my Net Engineers looking for dropped packets and other network anomalies, but they have not come back with anything yet.

       

      So my question to the community is: Has anybody else seen anything similar to this? And if so, what was the fix?

       

      Just as side note, I had posted about another issue with our SignUp stations where they are taking between 20-40 seconds to assign a PC or place in the queue. Pharos Tech support says the faulty PC issue could be related, but again, not much help coming from tech support.

        • Re: Faulty PCs in 9.0 R2
          Steven English Guide

          Adam Demeter,

           

          You don't have any settings on the NIC's that would result in the power being turned off to them do you?

          If you ping the system in question from the server does the workstation "wake up" and receive communications?

          What operating system(s) are in play here?

          What happens if you try to telnet from the server to the workstation - while it is faulty - using the network name of the PC in Pharos Administrator on port 28202 (e.g. "telnet comp_network_name 28202")?

           

          You should get a successful connection with some ASCII gibberish returned on that telnet test.

           

          Regards,

          Steven

          1 of 1 people found this helpful
            • Re: Faulty PCs in 9.0 R2
              Adam Demeter Scout

              Steven,

               

                    The NIC does have power management settings turned on. However they only kick in when the PC is in sleep mode (as per this article: https://support.microsoft.com/en-us/kb/2740020 ). The PCs never go asleep. The monitor is set to go off after 15min, but the PC does not sleep or hibernate. From the same Pharos server, I can remote desktop into the PC, with no issues, and ping the PC. I have not attempted to telnet. All the PCs are running Windows 7.

               

              Adam

                • Re: Faulty PCs in 9.0 R2
                  Steven English Guide

                  Adam Demeter,

                   

                  The telnet check should clarify whether or not there are any problems connecting to the SU client service on the workstation.  If the telnet attempt fails while the machines are faulty, then succeeds after you move the mouse, I would not expect the Pharos SU Server to successfully reach downstream and interact with the service either. At that point, diagnosis of why the client service is not listening (client service issue, NIC issue, etc.).

                   

                  Regards,

                  Steven

              • Re: Faulty PCs in 9.0 R2
                Adam Demeter Scout

                Steve, I'll give the telnet a try. But I may have yet another factor in the situation (not sure if this is related or not). So, we have 3 areas that use the SignUp process. One in a building across campus (building 14), and two in the same area (Library and Loggia). All of the areas are experiencing the slowness/timeout issue with SignUp. Building 14 doesn't seem to have as much of an issue as the Library and Loggia, but it still has slowness. I found out the other day, that the Loggia PCs are connected via Wi-Fi (not all of them, but out of 63 PCs, 25 are Wi-Fi ). Do you think this may be causing the issue? Again, this was not an issue with 8.4, but only since 9.0 R2 was installed.