NASLite Network Attached Storage

My old NASLite box finally died last week and yesterday I built a new box (Atom D525 processor, 2GB RAM).

Booted for first time last night and it got as far as checking the first (of five) drives. And for the last 12 hours it's been stuck at:
Checking storage filesystem (Disk 0).....Busy

I realise the disk checking process can take a long time, and they're big drives (1 or 1.5 TB), but am I right to assume that 12 hours is more than long enough to check a single disk, and there must be some underlying problem? If so, then any suggestions would be appreciated.

EDIT: I should add that there does appear to be disk activity.

12 hours sounds a long time - i have a 1tb and that can take an hour - the usual recommendation is to increase the amount of ram but thats usually on systems that have 128 or 256mb - i would have thought 2 gig was enough

i dont think the fact that there is more than one drive in the queue matters but it may - so it may be an idea to disconnect some of the others

other standard recommendations are to upgrade/reset bios and check all cables and connections

i would be patient but sooner or later you have to make a decision - if it is that it is just slow to disc check and you stop it you have to start all over again

not much help - sorry - others may know more but posting is slow as i think many of the regulars are busy with real life

Thanks for the speedy response. Given that there is disk activity, I'll try to be patient.

Well, we're now getting close to 24 hours, and it's still stuck at the same point (with the HDD activity LED still blinking madly).

My patience is running thin, and this can't be doing the hard drive any good. Do I just switch the thing off? Any further suggestions?

This suggests to me that that drive may well be goosed as an example my box had to do a file check the other week (it refused with 2 Gb ram but was ok with 6) it has a 13.7 TiB raid 5 array and eventually finished in arround 4 hours.

Your times do sound excessive me I would be tempted to end it disconnect that drive and try again with just the others connected and see how you get on.

Maybe this drive is what caused the demise of your previous box or that box has seriously corrupted the drive.

Other option is to fsck the drives using a live CD of anaother linux system (Ubuntu for example) not to sure how to go about this but I am sure someone will pipe in.

Hope this is of a little help.

Doug

i would stop it but before restarting anything with naslite boot the system with a linux distro - ubuntu or mint - both will boot from cd and allow you to look at that drive - depending on what you see you can make decisions from there

1 - you see nothing - then format it - that may not be good news if you want the data - you could try a recovery program - it might get the data back

2 - you see the files or some of them - copy them off and format it and then use it again in naslite

3 - you could try a disk check from linux - that bit i do not know how to do but know it can be done - google will tell you i am sure

4 - you cannot see anything and it will not format [unlikely] - turn it into a door stop

Thanks both for the advice. Decided to go the route of disconnecting all discs and then re-connecting and re-booting one at a time. Up to three and okay so far...holding my breath...

The drive hammering is not an issue for the life of the drive. there are drives that do nothing but random hits for their life and still run for years. there is no ware between the heads and the platters since they float on a cushion of air and there is no contact.

Mike

After many years of building and maintaining my own computers, I sometimes feel that I understand the processes involved. But mostly, I figure it's just white magic....

I spent many hours over the last couple of days trying to get my new setup to work. I could get two or three drives to work -- any two or three. But add all five, and the thing wouldn't even boot.

So today I got down to the desperate "try anything, even if it doesn't make sense" stuff. And now it's all working. To cut a long story short, I was using both onboard SATA (four ports) and a PCI SATA card for my drives. I found that, with a particular pair of onboard SATA ports, I could use any one of them, but not both. A repeatable, but mystifying situation. But at least it's all working now.

My only remaining problem is that my new relatively high-powered server has lower network speeds that my ancient box (I now have max 5MB/sec transfers on a 10/100 network). Also huge numbers or errors/dropped packets. But I'm still using v2.62, so I'll upgrade to the latest version before worrying too much about transfer speeds.

From time to time I have to make this statement.

NAS does NOT require a fast processor and allot of RAM. Look at any old Network Appliance box and you will see. P200 and it could easily sustain around 400Mbps around 1998. Your issues are because of IRQ sharing. I would bet that at least one or more of the SATA controllers and the NIC are using the same IRQ. Server hardware and servers are less effected by this. A fast new board and the latest processor are not likely to gain you much in the way of performance.

Mike

Thanks again for the help, Mike. I never expected a gain, I just didn't anticipate the loss. I did try varying the IRQ settings as a first step, but it didn't help. But I did wonder if this entry in the network status was a clue:
r8169: eth0: unknown chipset (mac_version = 1).

-- which is why I thought updating to v2.63 might be the next logical step.

Sounds like it is using a compatibility driver for the NIC.... Me like Intel and 3Com NICs personally and would recommend that you get one of them so you never have to look back.

Mike

Updating to v2.63 seems to have done the trick on network speeds (reads 7MB/s, writes about 9MB/s), so very pleased (although for the record I should add that I changed SATA mode from IDE to AHCI at the same time, so I don't know if that had an influence). NASLite still doesn't recognise the ethernet chipset, but it doesn't seem to be harming speeds. Thanks to Mike and the others for their help.

One thing: I don't understand the "dropped packets" entry below -- is it anything to worry about?
# RX packets:379154 errors:0 dropped:56002079 overruns:0 frame:0

And I wonder if I might finally share a last mystery about my new system? My system beep has disappeared! Or more correctly, was never there in the first place. My new system is silent on bootup. I've searched the BIOS for settings for system sounds, and can't find any. It's a small thing, but I'm used to listening for the system beeps when I reboot/shut NASLite. Have they done away with beeps on modern boards?

If it was me ... I don't like all those dropped packets. Back to mikeiver1's advice to get a new NIC.

Beeps: they are still there in software. But I also don't get beeps on an Intel Atom board (D945GCLF2). I just looked at the manual and it says:


(Minor Side Note: the "Server Message Log" shows ACPI messages with a board model missing the letter C.)

Georg -- thanks for the reply and apologies about the delay in my response. I've been away for a few days without access to internet -- always worth doing now and again.

Yes, I'd love to change the NIC (I've always used 3Com up to now). But my Atom board has just a single (PCI) expansion slot, currently used for an essential SATA card. So the built-in NIC is my only option.

Having said that -- a quick Google (should have done that before) suggests that having billions of dropped RX packets is not an unusual problem with Linux, and can be down to network settings in particular versions of the kernel. It doesn't seem to be a hardware issue. Anyone else experiencing this with v2.63?

And re the system beep -- I hadn't even thought about attaching speakers! But now I have, and that didn't help. Not a big deal.