SNAP and MS-Snipping-Tool problem

snap Problem Reports
#1

I’m a sysadmin at the Department of Geography at the University of Zurich and we use SNAP on our course room NUCs and on RDS for some courses.

We deploy SNAP as a PSADT package on the course infrastructure (NUCs) and on the RDS, but the problems also exist in a simple installation with the installer.

We use Windows10 on the NUCs and Windows Server 2022 for the RDS.
We are using SNAP 9.0.6.

There are currently 2 major issues:

1.) When SNAP is open and students use the Snipping Tool to take a screenshot, SNAP force-closes without any error. SNAP force-closes when students press the X on the Snipping Tool and discard the screenshot. I can sometimes reproduce this, but not always. WIN-SHIFT-S (snipping tool without GUI) avoids this problem.

2.) SNAP does not start. The splash screen appears but hangs. My guess is that the last time SNAP was closed, both processes (SNAP.exe and the ZULU Java task) were not closed properly. After using Task Manager to kill the hanging process, SNAP starts as expected. This may be related to 1.)

If I can provide any additional information, please let me know.

Thank you

#2

Dear @rothjoerg,
Thank you for reporting this. JIRA ticket SNAP-3573 created to track the issue.

#3

Dear @rothjoerg,

Please:

  1. upload SNAP’s logs of one of the affected machines. You can find it under %userprofile%\AppData\Roaming\SNAP\var\log.

  2. check Windows’ logs for errors, with special attention to graphic card errors.

#4

Attached my messages.log
messages.log (75.8 KB)

I can’t see any errors related to the graphic card in the Windows logs. It seems MS-Snipping Tool closes properly.

Regards,

Jörg

#5

[…] we use SNAP on our course room NUCs and on RDS for some courses.

We deploy SNAP as a PSADT package on the course infrastructure (NUCs)
and on the RDS, but the problems also exist in a simple installation
with the installer.

We use Windows10 on the NUCs and Windows Server 2022 for the RDS.

Given your infrastructure, and all the possible combinations:

  1. Running SNAP and the snipping tool locally on a NUC;

  2. Running SNAP and the snipping tool remotely on Windows Server 2002 and acessing it from a NUC via RDS;

  3. Running SNAP remotely on Windows Server 2002 and acessing it from a NUC via RDS while running the snipping tool locally on the NUC;

  4. …

it is very difficult to replicate the issue.

So, please help us help you: what is the simplest scenario where the issue manifests itself? Provide as much information as possible about this scenario.

Only if the scenario is within our capabilities of duplicating it, then we might be able to replicate the issue and therefore help you.

SNAP does not start. The splash screen appears but hangs. My guess is that
the last time SNAP was closed, both processes (SNAP.exe and the ZULU Java task)
were not closed properly. After using Task Manager to kill the hanging process,
SNAP starts as expected. This may be related to 1.)

Since SNAP’s process is still running, and since the Snipping Tool is the trigger for the issue, our current assumption is that this does not happen because of SNAP.

This is given further weight by the fact that there were issues with the Snipping Tool crashing on Windows 10 (search for KB5012599 and snip).

Are your systems up-to-date with their patches?

#6

With version 9.0.6 it the following steps are necessary to reproduce it (independent of OS and distribution).:

Launch Snap:
Open attached tiff
Bands => band_1 => Add Elevation Band
elevation => Open Image Window
Image => Zoom in or out

Launch Snipping Tool:
Mode => Rectangular Snip
Take Screenshot
Press X to close Snipping Tool-application and discard changes

With version 9.0.6, Snap crashes (not always but most of the time). With Snap 9.0.8 I cannot reproduce it at the moment.

Our systems are patched to the latest.

Because we work in a managed environment (with version 9.0.6), I have opened 2 more tickets.

Thank you,

Jörg

#7

I usuallly download a geotiff from this site:

I can’t upload my example tiff because it is too big.

But the problem should be reproducible with other data as well.