Making my Synology NAS calm down

I logged into my NAS

$ ssh

I then could identify what process was burdening the system:

$ top

And it P to sort on CPU usage. It quickly became appearant it was the face recognition software that was killing the ARMv8 processor.

According to the interweb, it should be sufficient to disable the face recognition album in Synology Photos. It was not. The NAS still ran and re-ran the face extraction process which made it close to unusable.

So, I decided to solve this the good old Unix way. I got a hold of the PID (well, I could have gotten that from top, but it changes some times, so I prefer using ps):

$ ps auxww | grep synofoto-face-extraction
29746 root      20   0   93.6m   2.4m 0.000 0.488   0:12.63 S /var/packages/SynologyPhotos/target/usr/sbin/synofoto-face-extraction

and tried to tell it nicely to stop:

$ sudo su -
# kill <pid>

When that didn't help, I used force:

# kill -9 <pid>

Now, onto the clever bit: I replaced the Synology binary with my old tiny program. To do this, I first took a backup of the original app:

# cd /var/packages/SynologyPhotos/target/usr/sbin/
# mv synofoto-task-center synofoto-task-center.orig

Then, I created my own synofoto-face-extraction:

# vi synofoto-face-extraction

and entered the smallest program I can think of:

#! /bin/sh
exit 0

That means it will run an sh process and return immediately with a successful exit code, 0. The only thing left to do now, was to make the command executable:

# chmod +x synofoto-face-extraction

With this hack in place, my NAS finally calmed down. All programs that want to call out to synofoto-face-extraction can do so as before. And since it returns 0, the programs using it are happy. Not only that, all Synology apps and web interfaces instantly became faster and easier to use too!

After months of CPU spikes no matter what the NAS was up, I can now appreciate a resource monitor speaks of peace:

Synology taking a breahter

Licensed under CC BY Creative Commons License ~ ✉ torstein.k.johansen @ gmail ~ 🐦 @torsteinkrause ~