[SOLVED] Disk seems to always spin

[Maxiride]

Running the latest fvdw firmware on my Lacie Cloudbox, however I'm noticing that the disk seems to always spin. The device is in my house corridor and almost everytime I pass by the disk is running with no direct activity (i'm not transferring nor accessing the console).

Is there some way to track what cause the device to "turn on"? Might it be the router that sends some ping every now and then? :scratch

[fvdw]

Things you could check

Is the spin down feature activated?
It relies on the standby feature of the harddisk itself. Maybe you should change inactivity time setting to another value and then back again to old value. Did you change your disk ? If so does the new one support standby command?

Is the noflushd daemon loaded?
You can check that by giving command ps -ef on the command line and see if noflushd is in the list

Which features are active?
Twonky is one of the candidates that can keep a disk busy when it updates it database. But should stop after some time dependent on the number of files in its database.

Clients in your network contacting the NAS?
Normally routers do not send a ping.

If the above doesn't give a solution then finding out what the cause is is not easy. In the past I made a version of noflushd to minitor inactivity and spin down commands and when disk wakes up but it will not tell what application accessed the disk. Need to check my archive if I still have it.

[fvdw]

Found it, you can run noflushd from command line and enable debug.
Just kill the running daemon and restart it from the command line like this

Code: Select all

/usr/sbin/noflushd -d -v -n 15 /dev/sda

Note with n -15 disk should stop after 15 minutes of inactivity, of course you can change that value to see quicker standby command action

command to list all options

Code: Select all

/usr/sbin/noflushd --help

[Maxiride]

Thank you very much for the guidance, I'll try to debug the issue soon =) :)

[Jocko]

Hi Maxride,

There are many cases which may explain why disk spindown fails.

The usual cases are if some p2p servers are running like transmission or mldonkey. (Need to check if resilio (btsync) may get the same behaviour).
Other cases are if you opened some ports on your router for servers like FTP or apache. In first case, if a remote user sniffs the port 21 (which is common), there will be some writing on ftp log file and then disk spindown will fail (same thing if a remote user tries to get an access on your web server)

Otherwise, I just found there is an issue with vnstat daemon (network monitoring). Currently, it saves its cached data every 10 minutes. So if you use a period value grant than 5 minutes with noflush, the disk will never spin down.

[Maxiride]

I found out that my SmartTV box was continuously refreshing the DLNA library like every 15 minutes, so no issue on the fvdw firmware but just a bad setup of mine =)

Thanks for help =)