I’ll add more explaining… It’s run only one time before you start the node, to setup the directories and etc. I updated my old guide also (linked in the first post).
yep. but how you represented it - they are run every time.
I don’t get what you mean… I run that command only once, when I setup the node. It’s copy-pasted from the offical docs.
Thank you so much! I had done most everything the same as you except configuring the yaml as I do most of my setting when creating the container.
Thanks again and I’ll post updates when I can…
Bad idea… The sysctl.conf is often overwritten when DSM updates. Better to just put the sysctl commands in a scheduled script on boot. That’ll survive all updates.
1 Like
Is already mentioned in my 6x edited post 
, but thanks for drawing attention to it for newcomers.
I tried to avoid the config also. I only modified some params there that will never change.
Hey again!
Thought I’d give y’all an update. I tried the things you listed and most definitely moving the logging off the data drive to a different volume and changing to the custom levels had the biggest impact. Also, adding the memory seems to have helped as well. I turned off the memory compression and that actually made things worse for my DS918. Again I appreciate all the help and example configs. Now with this large data influx writing to 2 nodes my 918 is running at about 80% continuously. When I see the volume size differ too much from the nodes reported size I’ll manually run file-walker. Odd that the NAS units have such a hard time with I have a couple of old Atom systems that are only running 50% with the same amount of ingress on 2 nodes as well as full logging. I guess Ubuntu is just better than DSM for this kinda stuff.
Yeah, I turned back on memory compression too. I didn’t saw any difference, but I didn’t tested to much.
The linux distro used by DSM is pretty old and lacks many features from newer distros.
1 Like