There’s been much discussion recently of used space calculation issues and I think my node has been affected as well, but in a different way. My node is still running 1.66.1 and all I did was run Windows Updates and restart the node last night. From before the update to after the update approximately 200GB of used space vanished, it’s not in the Trash. I haven’t had a chance to check on the actual used space on the disk yet.
2022-11-17T20:17:14.702Z INFO piecestore upload started {“Piece ID”: “JRPSJ3J5LBFPHXX4GW6UQP7QRKA7W2X2XS7US2BPWW5PPB7WKUSQ”, “Satellite ID”: “12EayRS2V1kEsWESU9QMRseFhdxYxKicsiFmxrsLZHeLUtdps3S”, “Action”: “PUT”, “Available Space”: 242335410944}
2022-11-17T20:17:14.871Z INFO piecestore uploaded {“Piece ID”: “JRPSJ3J5LBFPHXX4GW6UQP7QRKA7W2X2XS7US2BPWW5PPB7WKUSQ”, “Satellite ID”: “12EayRS2V1kEsWESU9QMRseFhdxYxKicsiFmxrsLZHeLUtdps3S”, “Action”: “PUT”, “Size”: 22016}
2022-11-17T20:17:14.986Z INFO piecestore upload started {“Piece ID”: “PAK56XDHC6UXS6GSH72PB7URDT5Z44Y4GJOFKPASSF5A5UBWOAWQ”, “Satellite ID”: “12EayRS2V1kEsWESU9QMRseFhdxYxKicsiFmxrsLZHeLUtdps3S”, “Action”: “PUT”, “Available Space”: 242335388416}
Node restarts and I change config.yaml from 15.5TB assigned to 16.0TB assigned.
2022-11-17T20:20:30.093Z INFO piecestore upload started {“Piece ID”: “SOABNDJ234Z7JGME2T4PX4JRGY2FVE6U6FI5BLPUIVY75J4FLTAA”, “Satellite ID”: “12EayRS2V1kEsWESU9QMRseFhdxYxKicsiFmxrsLZHeLUtdps3S”, “Action”: “PUT”, “Available Space”: 950130623488}
Suddenly the node now has 950GB of free space instead of the 740GB it should have! I increased the allocated space as I’m away for the weekend and didn’t want the node to fill.
14656876 MiB ~15.37TB is used, in the allocation of 16TB you should have 0.63TB free, unless your databases are not updated with the current info.
If you disabled a filewalker on start, it will run a hour later and should update your databases.
I believe the cache should be updated every hour, this flag should disable a filewalker only on start.
If you have a different experience - let me know.
I haven’t disabled the filewalker on start. I have limited the number of concurrent retain requests to 1.
The filewalker takes 10’s of days to run on my hardware, so I will let it run as long as possible and then check the databases at the next update to 1.67.3. However there are no errors in the log related to the databases.
The retain job/process completed and I now have 142GB in the Trash according to the dashboard. However ‘Used’ shows less but the actual folder size is more…
Please note, Windows uses binary measure items, but Storj software uses SI measure items, so what is Windows shows as MB is actually MiB.
15,078,493 MiB = 15,810,945.875968 MB = 15.810945875968 TB
Dashboard accounts as used blobs and trash folders. All other files in the data folder (749 MiB or 785.383424 MB or 0.785383424 TB) are not accounted, so 15.810945875968 TB - 0.785383424 TB = 15.025562451968 TB
