-
Notifications
You must be signed in to change notification settings - Fork 1
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Z-Way crashes on WebSocket push. #101
Comments
mhmm ... I think it happens after a specific count of vdev metrics changes I've observed this crash with apps (since yet I'm not able to force it):
Today I've added the Astronomy module from store (needs active Base Module) and after an hour (and lots of changes by the device created by this module) the server crashes with the same error please have also a look at the log file: but here is a different error:
Also I did some changes in AutomationController last months affecting ws.push() ... Z-Wave-Me/home-automation@08617c7#diff-1abed9a9f6d6a5c9957285958c12b3c0 Z-Wave-Me/home-automation@e997df8 Z-Wave-Me/home-automation@1f35aae Z-Wave-Me/home-automation@d28a6af Not sure if this matters ... |
now I've restarted the server multiple times during tests and v8 debugging on Notification
|
Another one ... looks a bit different:
|
Stucked after some time using v8 debugger (not forced via browsers tap ~ idle) and DeviceHistory Module: poll every 1 min .... it crashes about 50 min after restart ...
|
@PoltoS , @pieceofsummer : Could please somebody of you guys have a look at this issue?? RPi 3 with latest v2.2.5 - UZB (5.06)
He's running a lot of Maros' apps. I also observe regularly a killed z-way-server on my RPI B+ with UZB (5.16) v2.2.5+ ... There is also a german thread in forum.z-wave.me where a user has same problem with RPi B+, v2.2.5, RAZ 1 ... z-way-server is regularly killed after some hours ... we've questioned him about his apps in use Thx. |
Maybe unrelated: |
Here's the stack from the german guy:
@PoltoS , @pieceofsummer can you please give some feedback? |
We have made quite a lot of changes in 2.2.5, so the original issue seems to be irrelevant Please show me how to reproduce the proble.. |
Hmm how to reproduce - good question, think there are all different influences: The german user for example has this setup: Modul : RaZberry 1 Chip 500 Apps active on system: All users have only start gdb debugger and waited till z-way-server crashes ... What about OS or SD Card Class? Could this also influence such behaviour? |
FWIW: To solve the corruption problem, I first tried to use the syslog server on my Synology212+. Unfortunately the Syno could not cope with the data bombardment at zway startup. Next I changed the configuration such that all logs were written to USB stick. No corruption since. Don't remember when I started that, but definitely over 6 months ago. |
@PoltoS , @pieceofsummer , @aivs My log is now written to an USB stick, seems that this improves the behaviour on my RPi B+ (RaZ 2) ... Im my case SD card is permanently stressed by z-way-server startup during developement :) |
Since upgrading to 2.3.0 a couple of months ago, this has been happening to me many times a day and causes the server to exit. I am now running 2.3.3 and it still happens.
I do not have any apps besides the defaults. |
This is known and fixed on 2.3.1. Now it is 2.3.4 |
In 2.3.3 gdb traps it but continue should resume the process as it ignores this signal now. Please try and fun the real reason for the process to crash |
Do you mean that your crashes are due to bad alloc and SIG_PIPE is not the real reason? |
That's how it looks yes. |
Hardly repeatable, pretty rare. Happens in mongose:
SIGPIPE on send:
"man send" says:
The text was updated successfully, but these errors were encountered: