Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[koots]

My DNS-321 with 2 1TB drives (raid 1) has been working great since I got it about a month ago, but recently, I can't access the volume at certain times.  That is, when I try to browse to it from Explorer, it waits and waits and eventually my Windows 7 PC pops up a message saying that it can't access.  I can ping the drive just fine, and I can log into the admin panel via browser fine also, and the status and log shows nothing interesting-- this is during times when I can't access the volume via explorer.  But then sometimes the volume will strangely become accessible again, sometimes after an hour or so.  I've also tried rebooting my PC to see if that might help, but no.....

I don't want a data drive that isn't reliable, and this drive is all of a sudden is giving me headaches.  Please help.....

[koots]

wanted to also add that I've also tried other obvious things like restarting the unit, turning drive hibernation off.

[jamieburchell]

Are you using Bit torrent or any third party add-ons?
Which drives are you using?
Have you tried removing each drive, attaching them to a PC and running manufacturers diagnostics or viewing SMART data if that's an option?
Have you tried a factory reset of the unit? You won't loose your data doing any of these things, but you should have a backup just in case.
Are you running the latest firmware?
Does the problem happen if you connet a network cable directly between your NAS and computer?

Just a few things to try

[koots]

Are you using Bit torrent or any third party add-ons?
Which drives are you using?
Have you tried removing each drive, attaching them to a PC and running manufacturers diagnostics or viewing SMART data if that's an option?
Have you tried a factory reset of the unit? You won't loose your data doing any of these things, but you should have a backup just in case.
Are you running the latest firmware?
Does the problem happen if you connet a network cable directly between your NAS and computer?

Just a few things to try

Thanks--- these are sorta helpful, if I had a computer I could hook these drives up to..... I'm running the latest firmware (1.03).  I'm using 2 1TB Barracude 7200 drives.  I haven't tried connecting a cable directly to the computer though, but that's not really a solution.  I guess it would tell me that my network is somehow negatively influencing my NAS, but......

I've continued to try various things.  It turns out that my right drive failed the SMART test.  So I powered down and removed it, so I could back up my data.  But much to my surprise, many of the directories were missing(!!!)  So I panicked a little and tried inserting the other drive into the left slot, and it appears that the missing data is on this drive (thank god.)  But it's all very disturbing......

So I reassessed what I've seen so far---
1) DNS-321 worked great for a month with ext3 and raid 1 config.
2) I accidentally cut power to the drive without shutting down.
3) since (2), it's been acting flakey, like described--- it sits there and crunches most of the time with the fan running and sometimes I can't access the data volume, even though I can ping and login to the admin page. 
4) One of the drives has many directories *missing*. 

I've decided that I'm going to rebuild my volume from scratch, using ext2 instead of ext3 because I found this on http://en.wikipedia.org/wiki/Ext3:

No checksumming in journal
Ext3 does not do checksumming when writing to the journal. If barrier=1 is not enabled as a mount option (in /etc/fstab), and if the hardware is doing out-of-order write caching, one runs the risk of severe filesystem corruption during a crash.[31][32]
Consider the following scenario: If hard disk writes are done out-of-order (due to modern hard disks caching writes in order to amortize write speeds), it is likely that one will write a commit block of a transaction before the other relevant blocks are written. If a power failure or unrecoverable crash should occur before the other blocks get written, the system will have to be rebooted. Upon reboot, the file system will replay the log as normal, and replay the "winners" (transactions with a commit block, including the invalid transaction above which happened to be tagged with a valid commit block). The unfinished disk write above will thus proceed, but using corrupt journal data. The file system will thus mistakenly overwrite normal data with corrupt data while replaying the journal. There is a test program available to trigger the problematic behavior. If checksums had been used, where the blocks of the "fake winner" transaction were tagged with a mutual checksum, the file system could have known better and not replayed the corrupt data onto the disk. Journal checksumming has been added to ext4.[33]
Filesystems going through the device mapper interface (including software RAID and LVM implementations) may not support barriers, and will issue a warning if that mount option is used.[34][35] There are also some disks that do not properly implement the write cache flushing extension necessary for barriers to work, which causes a similar warning.[36] In these situations, where barriers are not supported or practical, reliable write ordering is possible by turning off the disk's write cache and using the data=journal mount option.[37] Turning off the disk's write cache may be required even when barriers are available. Applications like databases expect a call to fsync() will flush pending writes to disk, and the barrier implementation doesn't always clear the drive's write cache in response to that call.[38] There is also a potential issue with the barrier implementation related to error handling during events such as a drive failure[39]

So.... not sure if this applies.  I'm going to install fun_plug (http://wiki.dns323.info/howto:fun_plug) so I can checkout fstab. 

I'm not getting that loving feeling from my NAS though.  I'm also not sure how much extra protection I'm getting from raid 1 on this guy.  It seem that a simple power loss has sent my drive into the weeds........ maybe there is someone at D-link that finds this as disturbing as I do.

[go140point6]

Does the problem happen if you connet a network cable directly between your NAS and computer?

I wouldn't suggest connecting your NAS directly to your computer with a standard ethernet cable, you need what is called a cross-over cable.  It's just an ethernet cable with the wires literally crossed over.  Although if you have a "dumb" hub/switch you should get the same results, since there is nothing to configure.  However, a bad hub/switch can cause issues like described... I see it all the time at work.

To the OP, don't forget to make a backup (a REAL backup) of the date you don't want to lose.  RAID1 is not a disaster recovery method... it's only redundancy.  If data corrupts and gets mirrored, then a drive goes out... you end up with the corrupt data that was mirrored.

I have to say, I really like my DNS-321 for all its little faults.  No it's not perfect, but it was (and still is) very inexpensive compared to other options (although prices are coming down all over).  I've had mine for about a year and just upgraded my drives to 2x2TB RAID1.  I took the old 2x1TB drives and put them in my main workstation in a RAID0.  So in each case, I have 2TB of space.  I work off my RAID0 and then synchronize the two volumes (for additions and deletes) on a regular basis.  Of course, my REALLY important files (tax files, etc.) I back up to a USB 2 TB drive and also upload to my work.  And because I'm anal retentive  , I make a backup the important stuff to DVD once a year.  My DNS-321 then does what I bought it for, mainly as a media center storage unit streaming content to my home theater PC.  And for that, it works pretty good!

[jamieburchell]

I wouldn't suggest connecting your NAS directly to your computer with a standard ethernet cable, you need what is called a cross-over cable.

While this used to be true, modern network interfaces feature Auto-MDIX http://en.wikipedia.org/wiki/Medium_dependent_interface which auto detect the physical connection type, eliminating the need for a specific cable.

To the OP, you neglected to mention the sudden power disruption to the NAS. Data corruption is a possibility on any NAS and file system. It sounds like one disk at least needs rebuilding.

Make sure you have at least one other copy of your data.

The suggestion to connect directly was intended as a troubleshooting step rather than a perminant solution.

[koots]

Some updates---- I took a look at fstab and there was no barrier option, so I'm weary of ext3.... does anyone have any info or experiences with this?  If I can corrupt my data by losing power, as it appears that I did, this is important info to consider. 

I went ahead and restored factory defaults, reformatted the drives (ext2) and configured for 2 separate volumes instead of raid1.  Reading around on the web, it makes sense to me now that raid1 is best considered to be a drive failure strategy to minimize or eliminate downtime.  When I was considering buying my NAS, I thought of raid1 as a way to practically guarantee that I won't lose my data--- this is not the case as I learned. 

It's got me thinking about data backup strategies.  Right now I have an automated backup program on each of my computers that backs up to the NAS, so the data on my computers are backed up.  As long as the data on my NAS is not the only copy, I should be safe.  But what I really wanted was to host my music and video libraries on the NAS and not need to keep backup copies updated.  Right now I plan to backup the video and music data to the other volume on my NAS.  But I'm interested in people's thoughts on this--- and any good strategies people are using--- my needs are not unique!   

[go140point6]

It's got me thinking about data backup strategies.

To expand on what I wrote above a bit, my strategy is to work off my RAID0 data volume on my workstation (RAID0 just for performance really) and then I use the "synchronize" function in the WinSCP FTP client to move my changes (additions, deletes, and modifications) to the RAID1 volume on the NAS.  I also thought as you did when I first got my NAS (and used it that way for almost a year) before I started to get uneasy with my "strategy".  As most will see, my current method's weak link is me... if I forget, get lazy, whatever and don't manually press the synchronize button, then no backup.  I could automate this process but the software presents me with a list of what is changing and I like to review it before committing the changes.  And for the most part, the changes I want to commit to the NAS only come after I am doing some work (converting movies, music, etc.) so it usually is pretty organic for me to finish my work by synchronizing.

I still think RAID1 has a place on the NAS... I like the fact that if one of my drives goes bad, I won't lose the data on that volume.  Of course, if the data is corrupt on that volume, it doesn't do me much good.  I have my NAS plugged in to a small UPS which is supposed to help regulate power issues.  And as stated, I keep my REALLY important documents backed up in multiple places.  I took a look and what I consider really important amounts to only about 21 MB compressed.

I would be very sad if my movies, photos and music disappeared, but it wouldn't be the end of the world.  The most cost effective solution might be to simply get a large capacity external drive and copy everything to it as a "point in time" backup and stick it in the closet.  Then once a year, do it again.  Anything else, in my opinion, is just over kill but I guess it depends how much value you put on your data.  I'm sure some people have very elaborate backup methods (duplicate NAS units, one off-site, etc.) but at the cost of money and time, is it really worth it?

[jamieburchell]

I have a DNS-323 which has a USB port, so have a backup to external drives. You could run 2 volumes and schedule a task to copy data from one to the other.