SOTI Pulse Logo

Mobicontrol app crash - device disconnects

E
Eline_

Hi, we are having a long-lasting issue that at seemingly random moments, the Mobicontrol app crashes/restarts. We are running our own app in lockdown and single-app mode with Native activity suppression. 

The devices that we are using are Samsung Enterprise, running different Android versions (11,12,13). Currently having SOTI agent 15.4.3 Build 1054 or higher installed. Has been happening also with older versions (first thought it was because of agent version, but turns out upgrading didn't do the trick). 

We see that when the crash happens (our own app re-initializes), in SOTI the logs show 'device disconnected' even though the network is stable. 

2 years ago
SOTI MobiControl
ANSWERS
RC
Raymond Chan Diamond Contributor
2 years ago

Have you officially opened any support case(s) in the past?   You might be told to configure the devices so that various device-agent and system log files can be retrieved for further analysis to possibly determine the cause of the crash/restart.

Just stating that crashes/restarts happen at random moments is too vague.   How often will the Mobicontrol agent app crash on a device ON AVERAGE?  Every few hours, days, weeks or months?   If it is frequent enough,  it might also be practical to turn ob debug-mode so that more data can be captured via USB ADB port during the debugging process.  This of course requires someone from Soti support team and/or from your side with the right technical expertise.

E
Eline_
2 years ago

Hi Raymond, I have opened a support ticket, but unfortunately no luck there yet with the solution.

I understand that it does sound vague, but it is very difficult to pinpoint the situation when this happens, which is why I was curious if anyone else experiences these seemingly random crashes. Or if there are any reasons that could be thought of as to why a device would 'disconnect' from SOTI even though the network is ok.

It is unpredictable when it happens, and not reproducible, which is why using the USB ADB port for debugging hasn't been successful yet. 

I have read in other tickets about disconnects in Mobicontrol happening because of doze mode or other settings, which is why I was hoping someone recognizes this and has a solution. 

P
P
2 years ago

Hi, yes i have a case opened for this since March 2023. Not much progress as the issue random and provided logs (Zebra's RxLogger) did not reaveal root cause. Due to agent crash / reload, lockdown kills production application, so users loose thier work and need to relogin. Correlation is, when phone call is being made (so IP connectivity is unavailable during phone call and restored after the phone call is terminated):

I am_proc_died: [0,31703,net.soti.mobicontrol.androidwork,100,5]
E AndroidRuntime: FATAL EXCEPTION: DefaultDispatcher-worker-1
E AndroidRuntime: Process: net.soti.mobicontrol.androidwork, PID: 31703
E AndroidRuntime: java.lang.Exception: pulse message timeout! switch to disconnecting
E AndroidRuntime:      at net.soti.comm.l.G(SourceFile:151)
E AndroidRuntime:      at net.soti.comm.l$h.invokeSuspend(Unknown Source:11)
E AndroidRuntime:      at kotlin.coroutines.jvm.internal.a.resumeWith(SourceFile:12)
E AndroidRuntime:      at kotlinx.coroutines.internal.b0.W0(Unknown Source:6)
E AndroidRuntime:      at kotlinx.coroutines.a.resumeWith(SourceFile:16)
E AndroidRuntime:      at kotlin.coroutines.jvm.internal.a.resumeWith(SourceFile:48)
E AndroidRuntime:      at kotlinx.coroutines.z0.run(SourceFile:105)
E AndroidRuntime:      at kotlinx.coroutines.internal.l.run(SourceFile:13)
E AndroidRuntime:      at kotlinx.coroutines.scheduling.k.run(SourceFile:3)
E AndroidRuntime:      at kotlinx.coroutines.scheduling.a.u(SourceFile:1)
E AndroidRuntime:      at kotlinx.coroutines.scheduling.a$c.d(SourceFile:15)
E AndroidRuntime:      at kotlinx.coroutines.scheduling.a$c.n(SourceFile:29)
E AndroidRuntime:      at kotlinx.coroutines.scheduling.a$c.run(Unknown Source:0)
E AndroidRuntime:      Suppressed: kotlinx.coroutines.x0: [t2{Cancelling}@6690388, Dispatchers.IO]

WT
Warren T
2 years ago

Hi both

I recently had issues with Soti crashing randomly and when this would occur the entire lockdown menu would also become unavailable. It might be unrelated but I was able to fix this. Bit of a different set up and different devices but as long as it doesn't harm your set ups in any way I can't see it hurting to try:

  1. Add a whitelist for the default webview being used on your device, as you have mentioned these are zebra I am unsure which is the default for these devices as we currently don't use them. You can add the app package id for that either via Application Run controls in the lockdown or you can add them as hidden launch items for the configured lockdown.
  2. Clear browser/webview cache (not full data)
  3. Restart the agent/device.

After a bit of digging we found randomly that the agent would do a check for white/black listed apps and would for some reason treat the default webview used in the OS as a blacklisted app and this would then cause the crash.

Maybe try with one device as a test and see whether this resolves the issue. Possibly a long shot but anything better than nothing. This issue was causing our on street fleet of devices a nightmare.

EDIT: Just to add we have an HTML lockdown configured on our devices which uses webview, but the above did fix the issue permanently for us since, unsure whether yourselves are using an HTML lockdown.

J
JEMOD@SOTI
2 years ago

Hi Eline,

Thank you for posting on SOTI Pulse! 

It would be appreciated if you could grab a set of ADB logs when the crash occurs if possible. I would like to have a look into the device logs a few seconds before the event occurs to investigate this issue more in- depth. 

Kind regards,

Technical Support Specialist | SOTI | +1 905.624.9828 | SOTI.net lDiscussion Forum | Log a Case Online l Facebook l LinkedIn l Twitter 

Similar Discussions