Twonky wakes up NAS at midnight

[kreeture]

Hi all,

I have been trying to get to the bottom of my NAS coming out of standby several times when there is no network access to it. It seems to stay spun down when twonky is disabled but when enabled it always seems to spin the drives up at midnight with very little actual disc activity.

The media scanning is set to manual so it is not that, and the corncobs of the nas are set to run at another time so it is not those either.

Any pointers as to why this may happen with twonky?
I am using firmware 3.6.1, which contains twonky 6.0.38.

Cheers.

[schumaku]

HDD spin-down hype vs. opeational requirements.

The access might be intended, or it's unwanted by Twonky- however, in my opinion be happy that it does spin-up and work reliably.

Most likley, Twonky will not review or modify the Twonky 6 code anymore.

[kreeture]

Schumaku, i think you miss the point. Yes the nas works but no one, and I mean no one including the qnap support team have any clue as to why the drives spin up.
Now imho this is bad engineering, yes if there was any justifiable reason for the drives to spin up I could accept it but as far as I can see (and anyone else for that matter) there is no such concrete reason.
So unless anyone tells me specifically what is happening when the drives spin up then this should not happen!

The qnap support team were clueless as to why it is happening and told me to wipe the system and start from scatch. Which sounds like a stab in the dark and not what I would expect from competent engineers.

[sl1000]

as far as i know there is a crontab job running at around 3 am when twonkey is enabled. do a crontab -l to check for any jobs.

[Briain]

Hi

The crontab job clears the Twonky image cache and you can comment it or completely remove it. If you change some if the images embedded in their music files, they will only show up after you manually empty the cache, or rebuild the Twonky database, so my guess is that one advantage of that crontab job is that any replaced images would show up within 24 hours. Personally, I've moved the Twonky cache, increased its size limit (both done via RPC calls) and removed the crontab job to delete the contents at 3 AM as I've a lot of 800 pixel embedded images and want the cache preserved (so I don't have to build it up every day; that would slow the responsiveness of my control point). Whilst in there, I also changed the other crontab jobs to only occur once a week, and thus far, I've not noticed any deterioration in anything (but I mostly use it as a media server, so most other services are disabled).

Below is how I now have it set, and I'd be interested to know if anyone recommends I change any of the below tasks from weekly back to daily. As I say, I've not noticed any problems, but then I mostly only use it as a media server.

# m h dom m dow cmd
1 3 * * 3 /etc/init.d/idmap.sh dump
2 3 * * 3 /sbin/hwclock -s
3 3 * * 3 /sbin/vs_refresh
5 3 * * 3 /sbin/clean_reset_pwd
4 3 * * 3 /etc/init.d/backup_conf.sh

Bri

[kreeture]

Its not the cronjobs i know as I have modified those to run at different times.
These random wakeups are really really annoying.

[schumaku]

Before it was midnight as per the subject- now it's random?!?

Bad engineerig? The HDD spin-down is implemented in the was that the HDD are signelled to spin-down when the access to the physical storage is not required for a configured time, and to spin-up the HDD if the physical access is required again.

Anything beyond is your own personal assumption or based wrong expectations.

[kreeture]

Seemingly the spin up is at midnight and other times, I do not have the time to investigate constantly.
The point I was making is that no one seems to know what processes are accessing the drives, just giving best guesses. Wake ups still seem to occur for people with all the application services disabled, so go figure.

[schumaku]

No access physical storage access required for the configured time == spon-down. Physically storage requirement spin-up. Nothing to figure. This is what is implemented - and it's working impressingly well.

When you want your HDD spin down over a certain time with the idea to save energy or avoind noise - power-off the NAS by schedule.

[kreeture]

Yes but what is needing the storage causing the spin up?
I have been trying to figure this out now for months and no one has been able to help, hence my frustration.
I can accept the drives spinning up if I know what process is accesing the physical storage but the script and top do not point at any specific process.
Midnight here and the drives spin up again but no real disk activity.

[schumaku]

Obviously _something_ is requiring the storage to spin-up.

[kreeture]

Obviously.

[kreeture]

Just noticed it also comes out of standby at mid-day as well, anyone else seen this behaviour in a TS-410?

[P3R]

Just noticed it also comes out of standby at mid-day as well, anyone else seen this behaviour in a TS-410?

So a 12 hour interval. Do you happen to use that interval for internet time (ntp) updates? Or maybe the Qnap is a DHCP-client or server with a 24 hour lease time (a common default lease-time that would lead to refreshes every 12 hours)?

[kreeture]

Hi P3R

Ntp updates are disabled, the qnap has a static ip address so i would think it cant be a renew issue.
My router is the dhcp server with 24 hour lease time, but the qnap still wakes up even when disconnected from the network :/

Tearing my hair out trying to find out what is causing the spin ups and blkdevmonitor.sh script didnt give any leads.
So any other ideas would be much appreciated.