NASLite Network Attached Storage

This box has been up and running and unchanged for a year. It has eight drives in it and is only actually powered up maybe 1 day every couple of weeks.

v2.06 08-2007 - Powered up yesterday and it come up but refused to let my Mac access some of the eight drives. The Telnet session locked up while I was trying to get it to do a drive check. Ultimately hit the power button on the front which caused it to shutdown in an orderly manner.

Left it until today when I had time to look at it.

Powered up fine.

Tried to access using AFP from my iMac. First of all it told me that this device does not allow Guest Access. Then a couple of tries again it let me in. Now three of the 8 drives it reports as "The operation cannot be completed because the original item for Disk-1 cannot be completed". So it won't let me access Disk-0, 1 or 2.

If I access using AFP from my Mac Mini (both are running 10.5.2 OS X) I can access Disk-0 and Disk-1 but none of the other 6 Disks?

Bizzarre

Access from either the iMac or the mini using SMB fails with an error message "Connection Failed" when trying to login as default Guest.

Anyone any ideas on how to stabilise this NAS box?

Regards


Mike

Now it is allowing me into Disk-0 and Disk1 but not Disk-3 and Status

I am getting very worried about this...

Correction Can access Disk-0 and Disk-1 but not Disk-2 and Status, but it reports Disk-3 as available but has mislaid the 250GB of content.

Very odd,

The Syslog is very big and has complaints about hdd and hdc

Any suggestions for saving this NAS?

#
# May 10 13:53:28 user.warn kernel: end_request: I/O error, dev 16:01 (hdc), sector 357568560
# May 10 13:53:28 user.warn kernel: end_request: I/O error, dev 16:01 (hdc), sector 357568560
# May 10 13:53:28 user.warn kernel: end_request: I/O error, dev 16:01 (hdc), sector 357568560
# May 10 13:53:28 user.warn kernel: end_request: I/O error, dev 16:01 (hdc), sector 357568560
# May 10 13:53:28 user.warn kernel: end_request: I/O error, dev 16:01 (hdc), sector 357568576
# May 10 13:53:28 user.warn kernel: end_request: I/O error, dev 16:01 (hdc), sector 357564432
# May 10 13:53:28 user.warn kernel: end_request: I/O error, dev 16:01 (hdc), sector 0
# May 10 13:53:28 user.crit kernel: EXT2-fs error (device ide1(22,1)): ext2_write_inode: unable to read inode block - inode=22347780, block=44695554
# May 10 13:53:28 user.warn kernel: end_request: I/O error, dev 16:01 (hdc), sector 357568544
# May 10 13:53:28 user.warn kernel: end_request: I/O error, dev 16:01 (hdc), sector 357568552
# May 10 13:53:28 user.warn kernel: end_request: I/O error, dev 16:01 (hdc), sector 357564432
# May 10 13:53:28 user.warn kernel: end_request: I/O error, dev 16:01 (hdc), sector 0
# May 10 13:53:28 user.crit kernel: EXT2-fs error (device ide1(22,1)): ext2_write_inode: unable to read inode block - inode=22347779, block=44695554
# May 10 13:53:32 user.warn kernel: end_request: I/O error, dev 16:41 (hdd), sector 4248
# May 10 13:53:32 user.crit kernel: EXT3-fs error (device ide1(22,65)): ext3_readdir: directory #2 contains a hole at offset 0
# May 10 13:53:32 user.warn kernel: end_request: I/O error, dev 16:41 (hdd), sector 0
# May 10 13:53:32 user.warn kernel: end_request: I/O error, dev 16:41 (hdd), sector 144
# May 10 13:53:32 user.crit kernel: EXT3-fs error (device ide1(22,65)): ext3_get_inode_loc: unable to read inode block - inode=2, block=18
# May 10 13:53:32 user.warn kernel: end_request: I/O error, dev 16:41 (hdd), sector 0
# May 10 13:53:32 user.crit kernel: EXT3-fs error (device ide1(22,65)) in ext3_reserve_inode_write: IO failure
# May 10 13:53:32 user.warn kernel: end_request: I/O error, dev 16:41 (hdd), sector 0
# May 10 13:53:32 user.warn kernel: end_request: I/O error, dev 16:41 (hdd), sector 4248
# May 10 13:53:32 user.crit kernel: EXT3-fs error (device ide1(22,65)): ext3_readdir: directory #2 contains a hole at offset 0
# May 10 13:53:32 user.warn kernel: end_request: I/O error, dev 16:41 (hdd), sector 0
# May 10 13:53:32 user.warn kernel: end_request: I/O error, dev 16:41 (hdd), sector 4248
# May 10 13:53:32 user.crit kernel: EXT3-fs error (device ide1(22,65)): ext3_readdir: directory #2 contains a hole at offset 0
# May 10 13:53:32 user.warn kernel: end_request: I/O error, dev 16:41 (hdd), sector 0
# May 10 13:53:32 user.warn kernel: end_request: I/O error, dev 16:41 (hdd), sector 4248
# May 10 13:53:32 user.crit kernel: EXT3-fs error (device ide1(22,65)): ext3_readdir: directory #2 contains a hole at offset 0
# May 10 13:53:32 user.warn kernel: end_request: I/O error, dev 16:41 (hdd), sector 0
# May 10 13:53:32 user.warn kernel: end_request: I/O error, dev 16:41 (hdd), sector 4248
# May 10 13:53:32 user.crit kernel: EXT3-fs error (device ide1(22,65)): ext3_readdir: directory #2 contains a hole at offset 0
# May 10 13:53:32 user.warn kernel: end_request: I/O error, dev 16:41 (hdd), sector 0
# May 10 13:53:32 user.warn kernel: end_request: I/O error, dev 16:41 (hdd), sector 4248
# May 10 13:53:32 user.crit kernel: EXT3-fs error (device ide1(22,65)): ext3_readdir: directory #2 contains a hole at offset 0
# May 10 13:53:32 user.warn kernel: end_request: I/O error, dev 16:41 (hdd), sector 0
# May 10 13:53:33 user.warn kernel: end_request: I/O error, dev 16:41 (hdd), sector 7280

Are some of the disks on PCI adapter card, and could it be faulty?

How big is your power supply and what manufacturer is it?

Have you tried removing power / data cables from some of the drives to see if you can get a clean boot with the remaining drives?

The first 4 drives are on the mobo
The next 4 are on the PCI card

Jeantech 300w

On reboot Disk-2 and Disk-3 did not load at all and NasLite is declaring only 6 drives attached and AFP and smb are not functioning at all.

Also, NasLite is now reporting the Verify Code as INVALID for good measure.

This machine is under extremely low loading, basically it is an archive machine which is booted once every two weeks or so. I keep all my ready use video and Music Library on my 1TB Time Capsule. Every two weeks or so I move off old stuff to the NasLite Box. This way all the drives get a spin every couple of weeks.

Time for a Drobo I guess that will blow a hole in my toy budget.

Verify code invalidation would imply a change to hardware, so something onboard may have failed....

Disconnect all good drives just to be sure. Move the bad drives to a different position on the controllers and run a full check and repair with bad block and hope for the best. When you have lots of drives in your box, heat is your enemy and power is your friend, so a big PS and lots of fans are a must.

Thanks

It is all up an running again after another reboot.

I am looking at alternate ways of storing and accessing media, so hopefully this won't be a problem for much longer.

Thanks for the input all

Mike

It is unlikely that the errors in your syslog are the result of how you store or access media. Since both hdc and hdd appear to trigger IO errors, I'd suggest you take a closer look at the cable, connectors and controller. After verifying that all hardware is functioning fine, I'd strongly recommend that you run a complete check and repair to make sure the filesystems are healthy. If you are not sure that the hardware is working properly, then full check and repair may potentially cause more harm than good. In that case run a RO filesystem check for a look at the status before you decide on a next step.

In any case, keep an eye on the syslog for errors such as the ones you posted. Those are not normal and should be handled.

Thanks, appreciated.

I re-sat the ide cable and power leads and it all seems stable again. No idea how they came loose, if indeed they were the culprit.

I'll run a RO check on them.

Thanks again

Hot / cold cycles can cause IDE plugs to work their way out of the socket (eventually). It's not bad practice to check these plugs when cleaning out the inside of the case, as I'm sure we all do periodically.....

I actually do! I remove and replug them a few times and also give the plug on the cable a bit of a pull to make sure the connection is clean and tight. Gotta say I do love SATA for the simple and reliable data connector though, always hated the IDC ribbon connectors. I have even been known to assemble my cables with nolox to lock out corrosion.

Mike

Can recommend an ink eraser to (after observing static precautions) buff up the RAM and PCI edge contacts.

Solved a mysterious flakey kernel panic that cropped up.

The RAM was the culprit and merely reseating did not clean the contacts.