Repository navigation
Pool status code 5 shown as 'Unknown status: 5' (UGOS: 'Data organizing') #286
Replies: 3 comments
|
Update: pulled the full enum from the UGOS web UI itself. The Storage Manager bundle maps pool status with 0 Normal · 1 Syncing · 2 Warning · 3 Degraded · 4 Corrupted · 5 Data organizing · 6 Creating · 7 Deleting · 8 Repairing · 9 Rebuilding · 10 Replacing hard drive · 11 Adding hard drive · 12 RAID upgrading · 13 Upgrade · 14 Waiting for sync So "Unknown status: 5" = "Data organizing", and the current |
|
Thanks Jan @jvilhuber, much appreciated - and merged. :) /tom |
|
Addendum: Included in the beta of the next release. Beta installation see here. Let me know if it works for you. :) /tom |
Uh oh!
There was an error while loading. Please reload this page.
Summary
On a DH4300 Plus (UGOS), a storage pool's status sensor shows
Unknown status: 5while the UGOS web UI reports the pool state as "Data organizing". The pool-status code→label map informat_sensor_value()only covers codes0–3, so5falls through to theUnknown status: {raw}default fromformat_status_code().Environment
Normal, fanNormal, no fault present.Current mapping —
custom_components/ugreen/utils.pyObserved
sensor.ugreen_nas_pool_1_status=Unknown status: 5Normal→Unknown status: 5at exactly 00:00 (a scheduled background pool task starting), with every member disk, the fan, and all disk temperatures remaining normal throughout.So code
5corresponds to the pool's "Data organizing" background-maintenance state (benign, not a fault).Proposed fix
Add
5: "Data organizing"to the pool-status map. PR to follow.Note
Code
4for pool status is still unobserved here, so I've left it unmapped — only5is confirmed in this report.All reactions