Background: Since 2013 disks all Geli password decrypted using scripts. This problem never existed with the previous method, No API and bash scripts to do the work using geli attach and zpool import. Now obsolete.
Truenas 12.U6.1
SMB 1 clients.
Problem: In converting to ZFS and using the api to decrypt/mount/restart services breaks SMB. One disk is fine but not more.
Months ago I converted my first disk to ZFS encrypted with password and decrypted via a script using the API in a perl script. All worked fine until now that I have also converted other disks. What happens now in launching multiple scripts / instances of the API? it seems to confuse SMB so 'Workgroup' no longer appears on the network upon boot/decryption of multiple disks, needing to manually log in and stop/start SMB for it to show on the network.
The image below shows the setup, Init triggers the script ZFSDisks script which starts the *.pl scripts (sretalla perl script author).
Obviously it needs some kind of streamlining but I don't know what, on the bash script I tried putting 'sleep 5' between command but did nothing and I have no programming knowledge. Or could this be a bug that the API needs to confirm service restared before allowing an new instance of the same API.
Truenas 12.U6.1
SMB 1 clients.
Problem: In converting to ZFS and using the api to decrypt/mount/restart services breaks SMB. One disk is fine but not more.
Months ago I converted my first disk to ZFS encrypted with password and decrypted via a script using the API in a perl script. All worked fine until now that I have also converted other disks. What happens now in launching multiple scripts / instances of the API? it seems to confuse SMB so 'Workgroup' no longer appears on the network upon boot/decryption of multiple disks, needing to manually log in and stop/start SMB for it to show on the network.
The image below shows the setup, Init triggers the script ZFSDisks script which starts the *.pl scripts (sretalla perl script author).
Obviously it needs some kind of streamlining but I don't know what, on the bash script I tried putting 'sleep 5' between command but did nothing and I have no programming knowledge. Or could this be a bug that the API needs to confirm service restared before allowing an new instance of the same API.