Hi,
I would like to know if you can help me, I have just connected the D2600 to my DL380, with 4 disks in the D2600, at startup the 4 disks of the D2600 are automatically initialized to “sda, sdb, sdc and sdd”. Which is already replacing my disks and creating a launch error.
I would like to know if you have a solution so that they do not automatically initialize in sda,sdb,sdc,sdd but rather in sequence
When I unplug the D2600 and restart the DL380 everything returns to normal
You always can use UUID instead: How do I setup static mount via /etc/fstab for Linux? - Storj Docs
Hi @Egaria I saw that you then managed to connect the d2600 to the dl380. I hope my advice in PM has been helpful. I want to ask you a question instead, I see from the photo that you inserted 18TB disks in the d2600. Did you recognize them without problems? I didn’t go that far, I stopped at 10 TB per disk for d2600.
Hello, yes they were recognized without any problem 
Good morning,
I have a small question, I have my I/O which increases very high when storagenode is executed, as soon as I stop it the I/O goes down to an acceptable level, you have a solution to see what is happening pass ?
The solution is - keep it running.
The usage depends on customers, so any behaviour is normal.
This is VERY weird. How is you storage is connected?
In a DL380 caddy
20 characters
Yes… And is there a difference?
How so ?
20 characters
Please, elaborate… I do not know, what do you mean…
I don’t understand your question either, my disk is connected with a caddy inside a DL380
I do not have a Synology. So, what’s that mean that you have a
?
At the startup the storagenode software is running what is called a filewalker. Shortly, it is walking over the stored files to determine which files to delete and to count all the files and their sizes, and some filesystems aren’t very good at this (XFS for example).
This can take even hours to complete, then the I/O should reduce significantly.
Impact of this high startup I/O can be reduced by setting the lazy filewalker option to enabled in the configuration file. This then runs the filewalker as a separate process with lower I/O priority.
This best effort I/O however isn’t supported on older Kernel versions and certain filesystems. For ext4 I believe the support was added after Kernel 4.4.
The additional thing you can do to reduce the I/O is to move the SQLite databases to a separate drive, ideally a SSD. There is configuration option for this as well, but you have to move the files manually first before changing the option in the config.
I believe he is using HP DL380 and the drive is connected directly to a backplane as it is installed in one of the drive slots on the machine.
