core: 0.103.1 Failed to start after upgrade - Traceback errors and Core errors
Home Assistant release with the issue:
0.103.1Last working Home Assistant release (if known): 0.103.0
Operating environment (Hass.io/Docker/Windows/etc.):
Linux VMIntegration:
Description of problem: Frontend will not load… Service is running, just cannot log in. No changes were made prior to upgrade.
Did discover Cell phones via HA App so not sure if the old device name is the issue as I saw this was corrected in 0.103.1 ???
Log file attached: Log.txt
Didnt have the issue when upgrading to 0.103.0
—Restoring VM SnapShot to recover…
Problem-relevant configuration.yaml entries and (fill out even if it seems unimportant):
Traceback (if applicable):
● home-assistant@homeassistant.service - Home Assistant
Loaded: loaded (/etc/systemd/system/home-assistant@homeassistant.service; enabled; vendor preset: enabled)
Active: active (running) since Wed 2019-12-18 16:54:32 CST; 3min 9s ago
Main PID: 601 (hass)
Tasks: 29 (limit: 4915)
CGroup: /system.slice/system-home\x2dassistant.slice/home-assistant@homeassistant.service
└─601 /srv/homeassistant/bin/python3.7 /srv/homeassistant/bin/hass
Dec 18 16:57:39 HA hass[601]: 2019-12-18 16:57:39 ERROR (MainThread) [homeassistant.core] Error doing job: Task exception was never retrieved
Dec 18 16:57:39 HA hass[601]: Traceback (most recent call last):
Dec 18 16:57:39 HA hass[601]: File "/srv/homeassistant/lib/python3.7/site-packages/zigpy/device.py", line 184, in request
Dec 18 16:57:39 HA hass[601]: self.nwk, dst_ep, cluster
Dec 18 16:57:39 HA hass[601]: zigpy.exceptions.DeliveryError: [0xcaac:0:0x8000]: Message send failure
Dec 18 16:57:42 HA hass[601]: 2019-12-18 16:57:42 ERROR (MainThread) [homeassistant.core] Error doing job: Task exception was never retrieved
Dec 18 16:57:42 HA hass[601]: Traceback (most recent call last):
Dec 18 16:57:42 HA hass[601]: File "/srv/homeassistant/lib/python3.7/site-packages/zigpy/device.py", line 184, in request
Dec 18 16:57:42 HA hass[601]: self.nwk, dst_ep, cluster
Dec 18 16:57:42 HA hass[601]: zigpy.exceptions.DeliveryError: [0xcaac:0:0x8000]: Message send failure
homeassistant@HA:~$
Additional information:
About this issue
- Original URL
- State: closed
- Created 5 years ago
- Reactions: 20
- Comments: 32 (16 by maintainers)
Commits related to this issue
- 103.1 is broken home-assistant/home-assistant#30060 — committed to mdonoughe/home by mdonoughe 5 years ago
Sorry about this. 103.2 has been released. Thanks to @panctronic for findig the fix.
0.103.1 also stalls on start for me. Had to update
docker-compose.yaml to:image: "homeassistant/home-assistant:0.103.0"Chiming in to say 0.103.0 worked in docker, while 0.103.1 seems to be stalling during startup. It’s been stuck at 100% CPU for ~2-3 minutes now. It starts a listener for tcp/1400 (SONOS discovery iirc), but not for tcp/80 (my configuration for Http listener).
It could be an infinite loop, given the look-a-like memory leak
Logs show no fun stuff that wasn’t in 0.103.0.
EDIT: I’ve been unsucessful in acquiring a stackdump of the python process. I’ve tried using pyrasite to no avail. I’ll be reverting to 0.103.0 before my server cries help. It’s at 1GB memory and rising now, after 14 minutes runtime, with cpu never dropping below 100%.
@balloob is it possible to manually push the 0.103.2 to the docker hub? or do we have to wait until tomorrow?
+1, encountered the same startup issue with docker. Reverting to the 0.103.0 tag fixed it.
Please consider moving the
stabletag back to this release until the problem can be diagnosed.Same thing as ocalvo, no cloud-polling integrations even after reverting to 0.103.0:
Same on my side, after upgrading to 0.103.1 (docker), HA isn’t starting anymore. No error in my log, just HA using 100% CPU and memory increase…
The previous 0.103.0 version works fine.
Latest (0.103.2) docker image works OK.
It takes only 1h30m for a new release to be online. Almost all of them are done: https://dev.azure.com/home-assistant/Home Assistant/_build/results?buildId=20453&view=results
Depending on your timezone, that could be tomorrow though
Hang tight everyone; this is likely related to https://github.com/home-assistant/home-assistant/pull/29491 (released in 0.103.1). Will need @balloob to look into it once he returns.
@balloob Is this related to #29491 that was included in this release?
The cloud-based integrations are still broken with 103.2:
SmartThings:
Ring:
I find hard to beleive that these 3 integrations failed at the same time: Ecobee:
Smartthings
Ring:
After the rollback my ecobee, smartthings and ring integrations cannot longer connect because of invalid credentials.
https://github.com/home-assistant/home-assistant/commit/ec8ea022731b762481c2c88e1b14b4c9561c0b19#diff-5d2c9ff6d430c1e32f280e40a3dc275a fixes this problem
Seems like it’s getting stuck in a loop processing requirements:
When
KeyboardInterrupting/srv/homeassistant/bin/hass --debug -c .I added a bit of debug logging and it looks like
async_process_requirementsis being called repeatedly for name=http, requirements=aiohttp_cors.