hola a todos me pasa lo siguiente, hace unos días empecé a detectar un error en las métricas del almacenamiento donde no coincidía la cantidad entregada por el dashboard con respecto a lo real utilizado por mi disco HDD pero no le di importancia. Ahora mi nodo esta offline y revisando el log de errores me encontre lo siguiente:
Agradeceria mucho que me pudieran ayudar, aclarar que mi nodo esta montado en windows.
Senor, ¿usted habla inglés también? Por qué español no está mucho común aquí. Quiero te ayudar, más en inglés. Por qué mí curso español no está terminado.
You’ve got a problem with the space used. So what’s the file system and sector size?
You can get rid of the problem at least temporary (in my case this temporary is already >6 months) using:
storage2.monitor.minimum-disk-space: 1MiB
i understand, but, how can i see the file system and the sector size ?, when i put this “storage2.monitor.minimum-disk-space: 1MiB” on the config file at the moment to start the node i get this:
Seems like you made a mistake in the config file. This error is quite useless BTW, didn’t give any clue about the real problem. Did you change the config file, out did you just add this line to the file? In the latter case, you’ve got maybe two lines with the same command?
exactly, i was made a mistake but i could solve the issue with the line and my node have load again
i dont undestand why the node doesnt detect the correct storage size
the
“storage2.monitor.minimum-disk-space: 1MiB”
has to be at least 500GB,. its the minimum requirement.
here its NTFS.
comand line with admin rights:
fsutil fsinfo ntfsinfo c:
replace c with the partition letter
PS C:\Users\Dave> fsutil fsinfo ntfsinfo c:
NTFS Volumeseriennummer: 0x0474a76674a758e6
NTFS Version : 3.1
LFS Version : 2.0
Sektoren insgesamt : 3.513.659.391 ( 1,6 TB)
Cluster insgesamt : 439.207.423 ( 1,6 TB)
Freie Cluster : 214.771.926 (819,3 GB)
Reservierte Cluster insgesamt: 1.323.232 ( 5,0 GB)
Reserviert für Speicherreserve : 1.311.944 ( 5,0 GB)
Bytes pro Sektor : 512
Bytes pro physikalischem Sektor : 512
Bytes pro Cluster: 4096 (4 KB)
Bytes pro FileRecord-Segment : 1024
Clusters pro FileRecord-Segment : 0
MFT-gültige Datenlänge: 1,33 GB
MFT-Start-LCN : 0x00000000000c0000
MFT2-Start-LCN : 0x0000000000000002
MFT-Zonenstart : 0x00000000087a5860
MFT-Zonenende 0x00000000087b04e0
MFT-Zonengröße : 172,50 MB
Max. Max. Erweiterungsanzahl für Gerätekürzung: 512
Max. Byteanzahl für Gerätekürzung: 0xffffffff
Max. Erweiterungsanzahl für Volumekürzung: 62
Max. Byteanzahl für Volumekürzung: 0x40000000
Ressourcen-Manager-Bezeichner: FF3891DD-863A-11EA-BE0E-DDFE7EDD7825
Propably, you should set the node to full (allocate less than)3.7 tb
if clusterzize is 4k, propably check for
db errors, (move db to ssd/flashdrive)
Bloated logfile
high fragmentation: fix with ultradefrag,
forget decommissioned sats data: How To Forget Untrusted Satellites - #57 by daki82
fatal errors in the logfile.
full temp folder when node stopped (files older than 2 days)
filewalker problems in the log.
You really don’t seem to understand this configuration option. This is about remaining space on the used disk.
i try with move the db files to another directory and restart the node but the node remains the same
please check the log for fatal errors.
could be the 1min timeout error.
no, its not, its the minimum free disk space to be there to start the service.
its only for testing purposes, and should not be used.
(i double checked and translated it into my native language)
for setting this wrong, the min 10% free on disk requirement was massively exeeded.
in this stage the data gets much fragmented.
however, there is also an error with the bloom filter, leaving some unpayed peaces on the disk.
but this normaly not causes the node to terminate, so an other error is present.
there is an error with the (unfinished) filewalkers maybe. should be in the logs.
Also the 24gb MFT could be part of the problem. pointing to the timeout error.
No, it’s about to do not allow to run if there is less than that amount of free space. So the reverse. It’s an additional protection against running the Storagenode container with a broken mount point, where data would be stored in the mount point instead of the disk.
So it’s about remaining space, and not about a minimum space of at least 500GB.
And although I’m even running all my nodes with this option since over half a year, no node even came below 25GB remaining disk space with even less than 1% fragmentation.
Did you also change the folders in the configuration file? Otherwise, they will be recreated in the same folder.
You for sure had already a problem with the DBs, since your bandwidth stats and disk chart were already corrupted.
Ihre Muttersprache ähnelt sich sehr zu meiner Muttersprache. Aber, wenn sie etwas behaupten wäre es vielleicht auch klug die Suchfunktion zu treiben. Das Problem was sich erstmal im Log zeigt, wird öfter auf diese Weise gelöst. Seh mal The node doesn't recognize the used space after restart und Disk Space is less than requested - Bug? - #10 by baker
Ich würde sagen, dass es eigentlich ein STORJ-bug ist. Die Standart-Werte ist 500G, obwohl das auch das Minimum einer Knot-Größe ist. So, ich habe diese Funktion eigentlich kaltgestellt.
Aber, es gibt öffentlich mehrere Probleme, sehr gut zusammengefasst in deinem Post 
Also with my option, stated above?
Very additional, because there would be missing more things like the ‘storage-dir-verification’, config file and depending on configuration also the credentials. I would put this option up for deletion, actually. In the end I never have seen a real use, but now and then it gives just trouble.
Indeed, show you logs again please
