NAS or not?

I am not convinced that there isn't an issue with my QNAP now.

I have bought two HGST Deckstar's, initialised the first one, and then the drive disappeared, only to turn in an error code. Messed about reinserting into other bays and it spins back up, eventually? The initial SMART info has a warning so I am doing a full scan as I type which is clearly going to take hours. Is it ever straight forward with computers?!
 
The scan has come back with a bad retired block count. The HDD is in the same bay as the previous disk that had that same issue. I'm really thinking the issue is more the NAS than the drives. Guess I will need to remove the drive and run some tests away from the qnap from the mac.
 
Well it looks like one of my new HGST's is duff, I must have found one of the 1% failures. On the plus side, at least I haven't got to send my QNAP back to Belgium for repair.
 
What are the sustained transfer speeds like in comparison?
Not formally but, given the 4% difference in maximum usage as well, there's a definite difference in transfer time for the same large file.

Disk, OS, Network stack. It's hard to say what's causing that network IO. It's intriguing. I'd SSH onto the NAS and see if I could install/run iotop/iftop in order to garner more idea as to what's going on.
I've only really analysed write speeds to my server and I get a good solid 118 megabytes/second with desktop - which slows down to about 80 megabytes/second after twenty or so minutes - presumably that's when the in-memory cache fills up. It's absolutely no comparison to your set-up as the server is Xeon powered with a ZFS filing system.
I'm convinced it's an interaction between disk and OS as I get a similar pattern with just two HGST drives in RAID1 in a different Asustor NAS. However, I won't be doing any more on it as Asustor has reproduced the problem and is investigating. :)
 
Last edited:
Well it looks like one of my new HGST's is duff, I must have found one of the 1% failures. On the plus side, at least I haven't got to send my QNAP back to Belgium for repair.
Unfortunate, but these things happen. Good that you found the root cause.
 
Unfortunate, but these things happen. Good that you found the root cause.

Seems to have been the issue. Sent it back yesterday, so hopefully get the replacement early next week. Everything else seems to be playing ball now. The USB 3 enclosure does seem snappy enough for accessing RAW files in LR etc and happy that I have stripped all the files off my mac HDD leaving just the OS. Took a fair while to sync everything to the QNAP but now done it will be a little and often.
 
I had a colleague order six disks simultaneously from the same supplier and 3 were dead on arrival.

Worth making the effort to check the disks first - in Linux I'd run the extended SMART test with smartctl and then run badblocks which will test a formatted partition for any blocks that cannot be read. It also has a write test mode, but that will destroy any data that is on the disk.

It's not to tricky to get your head around, if your the type of person who doesn't mind using the command line. You can run them from a live USB, e.g. Ubuntu, although you will probably need to install them first.
 
If you can afford to not have your material at "instant" access - could cloud storage be an option for you?

For me I use a simple ReadyNAS to store my material in the house as a backup. However, this is again all backed up to a cloud. (Paranoid).

The cloud is a lot slower, but if you don't need regular access - it could work. ie I back up everything to my NAS after every shoot, then back up to the cloud overnight.
It can take a while, depending on size of transfer and your connection speed - but its workable.

I'd probably consider it as a full time option for myself if I didn't have the NAS already - since its just backup and I dont use the NAS drive as a lightroom extension.
 
Back
Top