deconz-rest-plugin: deCONZ hangs for 5 minutes during startup
Not sure what’s happening here. Shortly after starting deCONZ v2.05.49, it freezes completely for ~5 minutes. The GUI doesn’t respond; when minimising and maximising the GUI, it doesn’t even redraw. The log is quiet during this time, but afterwards, the messages appear with a timestamp from during the freeze (note the difference between syslog and deCONZ time:
Dec 09 16:19:51 pi2 deCONZ[13372]: 16:19:51:307 Node: id: 58, 84:18:26:00:00:0b:ac:51 (0x84182600000BAC51) scene: -962.057463, -950.942537
Dec 09 16:19:51 pi2 deCONZ[13372]: 16:19:51:313 Node: id: 59, 00:17:88:01:02:74:ba:72 (0x001788010274BA72) scene: -1200.000000, -100.000000
Dec 09 16:24:19 pi2 deCONZ[13372]: 16:19:51:323 Node: id: 60, 00:17:88:01:10:49:55:03 (0x0017880110495503) scene: 546.000000, -544.000000
Dec 09 16:24:19 pi2 deCONZ[13372]: 16:19:51:333 Node: id: 61, 00:17:88:01:02:00:ac:7c (0x001788010200AC7C) scene: -1437.000000, -394.339920
After the GUI coordinates have been loaded, I see a PROTO: CRC error message:
Dec 09 16:24:19 pi2 deCONZ[13372]: 16:19:51:833 Node: id: 99, 00:15:8d:00:02:39:14:5e (0x00158D000239145E) scene: 546.233293, -373.441510
Dec 09 16:24:19 pi2 deCONZ[13372]: 16:19:51:881 Node: id: 100, 00:15:8d:00:02:4a:70:18 (0x00158D00024A7018) scene: -1346.989853, -508.468633
Dec 09 16:24:19 pi2 deCONZ[13372]: PROTO: CRC error
Dec 09 16:24:19 pi2 deCONZ[13372]: 16:24:19:032 DB save zll database items 0x000008C3
Dec 09 16:24:19 pi2 deCONZ[13372]: 16:24:19:228 DB saved in 196 ms
I’m also seeing complaints about the NWK update ID:
Dec 09 16:24:26 pi2 deCONZ[13372]: 16:24:26:894 CTRL [00:15:8d:00:02:4a:70:18] found reference in zll.db
Dec 09 16:24:26 pi2 deCONZ[13372]: 16:24:26:896 Beacon src: 0x51A2 ch: 25 updateId: 10
Dec 09 16:24:26 pi2 deCONZ[13372]: 16:24:26:896 * node has higher updateId should be 0, TODO handle
Dec 09 16:24:26 pi2 deCONZ[13372]: 16:24:26:896 Beacon src: 0x7E36 ch: 25 updateId: 10
Dec 09 16:24:26 pi2 deCONZ[13372]: 16:24:26:896 * node has higher updateId should be 0, TODO handle
Dec 09 16:24:26 pi2 deCONZ[13372]: 16:24:26:896 Beacon src: 0x4855 ch: 25 updateId: 10
The Network Settings in the GUI shows 10.
It seems like deCONZ is running normally, but:
$ ph get /groups/0
ph get: error: 3 resource, /groups/0, not available
Attached the log: log.gz
About this issue
- Original URL
- State: closed
- Created 6 years ago
- Comments: 18 (8 by maintainers)
Commits related to this issue
- Prevent device descriptor duplicates in database Issue: https://github.com/dresden-elektronik/deconz-rest-plugin/issues/1018 — committed to dresden-elektronik/deconz-rest-plugin by manup 6 years ago
- Delete duplicates in device_descriptors table Issue https://github.com/dresden-elektronik/deconz-rest-plugin/issues/1018 — committed to dresden-elektronik/deconz-rest-plugin by manup 6 years ago
No noticeable delay during startup, even with the full database. The database is looking good - no more duplicates in the
device_descriptorstable.After the cleanup, I shutdown deCONZ and did a manual VACUUM. This reduced the database file from 848,896 bytes to 306,176 bytes.