All users Freeze while working in OE

Configuration

- Sage 300 SP2 (and upgraded to SP3 no difference)

- remote services configuration 2012 R2

- approximately 5 active users processing daily orders and cash transactions in OE

- 15 users overall access sage

Frequently (5 or 6 times a day) one user will freeze and when that happens all users working in OE AR receipt batches and some IC inquiries or bank rec will also freeze.  All users remain frozen until we disconnect the original user that first got frozen and then everyone starts working again.   If a users is not actively working without any screens open when the freeze happens they are able to move around successfully in AP or GL but the minute the access a AR receipts batch and most things we have tested in OE they will also freeze.  

Has anyone seen this before and have ideas for resolution.   

Based on my monitoring till now i strongly believe it is related to when the OE people choose to record the payment with the ship all function of an order and access a cash batch.  In this environment we have 3 locations that do this and they all access a different cash batch (one for each location) which shouldn't cause a record locking issue but i belive it is related.   When we load bank services and start clearing deposits we can almost always bring on the lock/freeze.  The lock/freeze can happen first thing in the morning when the only people that are there are the OE sales people the rest of the accounting staff doesn't' come in till later so we know its not load related.  

  • One thing that causes poor performance is if you have a lot of transaction history in bank services, this is a known problem. The solution is clearing bank services history. I'm leaning this way since you say this happens when users access cash batches.

  • i did remove bank services history and there hasn't been any real Difference.  I am now getting more feedback from the users.  It appears that they quite frequently get a UI error, they close it and just try again and everything works.  Today when running a day end we received errors indicating that there was a record lock with EFT and it couldn't continue and to try again later.  I went to current users to see who might be in EFT.  It turned out only 3 users were listed on the lanpak tab.  But we knew that there was more like 10 or 12 users in sage.  We isolated a single computer that was not listed on the lanpak users screen and that computer could move around open ar customers bring up customers – account activity etc.   I would click on open company check the userid, confirm it was the user I was tracking, attach again.  Still not in lanpak users.  I exited sage, exited my RDP connection to the server and re-established my remote connection and into sage and then that user was on the lanpak users list.  If I did open company and go out and attach to sage as a different user the user was removed from lanpak and listed as the new user everything was as expected.   Users continued to have ongoing record lock errors for approximately 2 or 3 hours.  We forced all users to go out completely back in so their names were all in the lanpak list.  While I was watching this list – right before my eyes. There was a blink of the screen, like the refresh when it looks for new users, and everyone was gone completely from the list and I was frozen.  I went to task manager forced a close, went back in and the only user in lanpak was me.  All users again had to go all the way out and back in they re-registered in the lanpak list and now have been working fairly successfully over the last 2 hours. 

    Does anyone have any ideas what may be going on.  Is the list of lapak user not reliable and others have seen that this means nothing and I shouldn’t focus my attention on this.  Or is this an indicator that Sage has lost track of a user that could be locking files and is causing all these strange conflicts and record locks.

     

    Does anyone know why users show up on the “other” tab.  This customer has many users on the Other tab and I have never seen that before.  I have only ever seen users listed on the lanpak tab.

  • in reply to ValerieH

    I haven't seen the lanpak list do that before. It really sounds like some sort of network disconnect.  I don't think I saw it above - what version of Sage are you running?  Are the applications on the same server as the RDP server?  Is the configuration of the Path environment variable the same as how the registry is pointing to the Sage programs? (both UNC or both drive letter or ...).

    Is any of this within a virtualized environment?

    Is anti-virus running on this server?

    Is the server used for anything else?

    Do you have non-Sage applications accessing the Sage data (third party applications, for example)?

    The LanPak.bin file in the shared data\SITE folder is where the list of users will be.  Semaphor.bin, also in SITE, governs locking of resources.  If you have something that is trying to update or gain control of these files, halt that process.  There's no point to backing up semaphor.bin - if it doesn't exist then the first user into Sage will cause it to be re-created.  I think it is the same for lanpak.bin, but I'm not certain.  But if there is an issue with different workstations not interacting with the files in the SITE folder properly then that will cause locking and concurrency issues.

  • in reply to ValerieH

    Question: Are the users signing in using the workstation from their own PC OR are they first of all signing into a server with RDP that has the workstation installed.

    If the above is the case, then this is a known problem with windows 2012 R2 servers.  You have to have the programs also installed on the workstation RDP server as well. (I have upgraded many clients from windows 2008 to 2012 R2 and I get this problem where the WS RDP server loses the sage connection to the main application server - so I then installed the programs on the RDP server and the problem goes away)

    If this is not the case, then I have never seen this from PC workstations.

  • in reply to Django

    sage 2016 SP3,  Apps are on same server as RDP, env path and reg are the same both drive letter, yes virtualized enviroment, no anti virus (have shut down still issues), Server does also provide MS office apps, no non-sage apps, 

    you did put me on to something though with your explanation of the lanpak and semaphor file.  I was not aware of it as it was a bit disgused through the share but it turns out that the shared files (licenses site company and user) were on a different server from program files.  both local servers to each other so shouldn't matter.  But I made that change last night to place them on the same server, i did leave lanpak and semiaphore file behind (new ones created)   this morning everyone is in sage and they are correctly listed on lanpak tab and NOT on the other.  So I am hopeful.

  • in reply to IDLSystems

    they remote into a server and run the program files from the server they are remoted into.  But interesting to hear you have seen the issue i am seeing before and was related to a delay between program files and access to shared files so that would leave me to believe that what i posted above may be the right direction.

  • in reply to ValerieH

    Is it anything to do with the Signon Manager?  I am having weird locking issues and periodically it references Signon Manager so I don't know if that is something to look at.

  • in reply to ValerieH

    for anyone following this incident.  The change i made to have the lanpack.bin and semephor file on the same server has in fact appeared to resolve the problem.  Those two servers are local to each other in the same hosting facility which in my mind does not make sense but I can't argue with results.   

  • in reply to ValerieH

    They should never have been on separate servers in the first place, there should only be one SITE folder.

  • in reply to ValerieH

    Ditto... you should always only have one SITE folder. No wonder you were having issues.

    You are also violating Sage licensing if you have multiple SITE folders for the same installation.

  • in reply to 49153

    not sure what your saying.  I always only had one site folder.  It just didn't reside on the the same physical server as the program files.  

  • in reply to ValerieH

    I think when you described that you moved lanpack.bin and semaphor.bin to the same server it sounded odd.  Did you change the SharedData folder in the registry to be on the same server as the ProgramFiles folder in the registry? Or did you have something more complicated in your setup?

  • in reply to ValerieH

    You said LANPAK.BIN and SEMAPHOR.BIN were on different servers.