OI 9.4 app runs on Server not on Workstatiom (OpenInsight 32-bit)
At 22 MAY 2020 11:37:50AM Mark Glicksman wrote:
My OI 9.4app runs perfectly when run on the server, but crashes when run on a workstation.
Windows Server 2016
LH service lh47nul
netwotk driver lh 4.7.2
At 22 MAY 2020 11:40AM Mark Glicksman wrote:
Workstation is Windows 10 Enterprise
At 22 MAY 2020 12:08PM bshumsky wrote:
Workstation is Windows 10 Enterprise
Hi, Mark. Can you give some more details of what you mean by "crashes"? Does it start up at all? Does it give any error messages? Has this been working on this workstation before, and just stopped? Does it work on other workstations, just not this one?
Thanks,
- Bryan Shumsky
At 22 MAY 2020 02:26PM Mark Glicksman wrote:
Thanks for your help Bryan. Yes, it does start, but OI just disappears, often when it's trying to print to a file.
At 22 MAY 2020 02:34PM bshumsky wrote:
Thanks for your help Bryan. Yes, it does start, but OI just disappears, often when it's trying to print to a file.
Hi, Mark. I would try re-running the clientsetup.exe found in the OI folder. Please watch very carefully during the clientsetup process to make sure that there are no errors reported during the installing of the various bits n' pieces, and then see if that makes it more stable.
Hope that helps,
- Bryan Shumsky
At 26 MAY 2020 10:28AM Mark Glicksman wrote:
Bryan, I did run Client Setup. I will try it again, to make sure I don't see any error messages. Thanks.
At 27 MAY 2020 02:48PM Mark Glicksman wrote:
I did try running Client Setup again and didn't notice any error messages. I just ran into the same problem on another similar installation. Do you have any further suggestions?
At 27 MAY 2020 02:59PM bshumsky wrote:
I did try running Client Setup again and didn't notice any error messages. I just ran into the same problem on another similar installation. Do you have any further suggestions?
Hi, Mark. Sorry to hear that clientsetup didn't resolve the issue.
The fact that OI starts up and then just "disappears" (without any error messages about crashing, etc.) is very odd. I might suspect some kind of antivirus or something that's shutting you down, or maybe something that OI is running that causes it to terminate.
When you say you have the problem "on another similar installation", you mean another system entirely with a different server and workstations, or on a different workstation talking to the same server?
Are there multiple workstations talking to this (or these) servers? Do all the workstations show the same behavior?
Once OI crashes, can you restart it? Does it start back up again and crash again, or does it always/sometimes run normally?
- Bryan Shumsky
At 27 MAY 2020 05:19PM Mark Glicksman wrote:
Bryan, there are 2 other workstations, and they work normally with no problem. The program can be restarted. But, it always crashes again. When I ran the app in developer mode (using /dv), it ran for a longer time, but eventually also crashed.
At 27 MAY 2020 07:35PM Matthew Crozier wrote:
I've seen that problem (OI 9x just disappearing) when there are multiple concurrent sessions running without a proper connection to the Linear Hash service. If it's possible, try testing affected workstations by calling the NTSserial() function - field <2> should return the serial number of the LH service. If it returns 'N/A' instead, then there isn't actually a connection.
HTH, M@
At 27 MAY 2020 10:10PM bshumsky wrote:
Bryan, there are 2 other workstations, and they work normally with no problem. The program can be restarted. But, it always crashes again. When I ran the app in developer mode (using /dv), it ran for a longer time, but eventually also crashed.
Hi, Mark. I vaguely remember another scenario that produced results like this, and it had to do with how the user had indexing set up on their system. In your OI on the "crashing" workstation, what do you have the "delay before Indexing" set to? I believe you can find this in the Database Environment Settings (accessed via the Database Manager in 9.4.x, if I recall correctly…)
Thanks,
- Bryan Shumsky
At 28 MAY 2020 10:59AM Mark Glicksman wrote:
In both cases, "delay before indexing" is set to 60.
At 28 MAY 2020 03:10PM bshumsky wrote:
In both cases, "delay before indexing" is set to 60.
Can you try setting the delay before indexing to 0 for these workstations to see if it makes any difference in the crashing/disappearing? You can still have "update before query" checked, for example, so that the users will still have the latest results.
Thanks,
- Bryan Shumsky
At 28 MAY 2020 03:12PM bshumsky wrote:
In both cases, "delay before indexing" is set to 60.
Another question - do you use mapped drives in your application? If so, could you replace any mapped drive references to use UNC names instead? Sometimes the "resolution" of mapped drive names causes memory issues…
- Bryan Shumsky
At 01 JUN 2020 04:05PM Brad Bishop wrote:
Make sure Windows Data Execution Prevention on the client is set to Windows programs only.
At 09 JUN 2020 03:34PM Mark Glicksman wrote:
Bryan, it was using a mapped drive. I changed it to UNC, but that didn't solve the problem. I have isolated it to where the app attempts to print a text file using OSBWrite. That's when it crashes. This causes no problems for other installations for other clients.
At 09 JUN 2020 03:37PM Mark Glicksman wrote:
Brad, I am not able to access the DEP settings. It's possible that this client has not granted me permission.
At 09 JUN 2020 03:37PM bshumsky wrote:
Bryan, it was using a mapped drive. I changed it to UNC, but that didn't solve the problem. I have isolated it to where the app attempts to print a text file using OSBWrite. That's when it crashes. This causes no problems for other installations for other clients.
Hi, Mark. Where is the OSBWrite trying to write to? Is it somewhere over the network (via a mapped drive or UNC)? Is it to the local machine? Does the location you're trying to write to actually exist?
If it does exist, then I suspect there's some permissions issue perhaps going on…?
- Bryan Shumsky
At 09 JUN 2020 03:48PM Mark Glicksman wrote:
Bryan, it's trying to write to the local C: drive. Yes, the files already exist. The strange thing is the app prints successfully to other files in another folder on the C: drive. Perhaps it's a permissions issue with the folder it's trying to write to when it crashes?
At 17 JUN 2020 03:54PM Mark Glicksman wrote:
Somehow the problem resolved. I don't know if it was because of one of my actions or if the problem simply resolved itself. Anyway, thanks for your help.
Another sit is reporting sudden shutdowns. I'll post that in a new thread.
Best wishes,
Mark