Snap 1000 Error - no access

[microsvc]

I recently installed a larger drive in my 1000 (320gB) and it worked well. Over the past few days, it became unaccessible. Here's a few lines from my logs:

E File System Check : FSCK fatal error = 4 Disk 10000 7/24/2010 1:56:04 PM
E File System Check : Cannot Read: Blk 82304 Disk 10000 7/24/2010 1:56:04 PM
I File System Check : ** Phase 1 - Check blocks and sizes Disk 10000 7/24/2010 1:55:23 PM
I File System Check : partition is clean. Disk 10000 7/24/2010 1:55:22 PM
I File System Check : Executing fsck /dev/ride0a /force Disk 10000 7/24/2010 1:55:22 PM


Does fatal error 4 mean the drive has failed? The disk activity LED is on a lot.

Thanks. ~Bob

[rpmurray]

It doesn't mean the drive has failed, just that some sectors on the drive cannot be read (which is almost as bad). As far as I know, the Snap OS does not appear to try to map around bad sectors, so essentially your data is now hosed. From what I've read here over the years, you have two options:

1) If you can still log in, you could try setting the option to fix fatal errors and see if it does anything. This is a long shot and I've found that it generally doesn't work most of the time.

2) If you really need the data, then you'll need to send it to a data recovery place and have them recover it for you.

There are some folks on this forum that have suggested using SpinRite to see if it can fix the problem, but this has never worked for me. SpinRite is useful to test drives before installing them to see if they have any bad sectors.

If you've got a backup of your data; get a new drive, possibly run it through SpinRite to test for bad sectors, and then install it and restore your data. Once a drive starts getting read and/or write errors, it's useless in a Snap because the OS is not designed to handle them; to the OS it's like running into a brick wall.

[bitor]

Hi microsvc,
When you say unaccessible are you talking about the webgui share folders? Have you tried to ftp in to the unit?
There are limitations to the SnapOS as far as files and folder maxs and file size.
Have you tried to reset the unit? Also, I would check the drive with HDD Regenerator. I find this software to be the best. If you think it might have a problem. Also, check to make sure the fan is running and not going bad. I believe the fan kicks on at 110F from my memory.

bitor

The light coming on a lot could be from disk swaping more so than normal because the lack of RAM memory(32MB) and large hard drive. When you install larger HD on these units I highly recommend getting the most memory possible on the hard drive's PCB board.

Originally Posted by microsvc

I recently installed a larger drive in my 1000 (320gB) and it worked well. Over the past few days, it became unaccessible. Here's a few lines from my logs:

E File System Check : FSCK fatal error = 4 Disk 10000 7/24/2010 1:56:04 PM
E File System Check : Cannot Read: Blk 82304 Disk 10000 7/24/2010 1:56:04 PM
I File System Check : ** Phase 1 - Check blocks and sizes Disk 10000 7/24/2010 1:55:23 PM
I File System Check : partition is clean. Disk 10000 7/24/2010 1:55:22 PM
I File System Check : Executing fsck /dev/ride0a /force Disk 10000 7/24/2010 1:55:22 PM


Does fatal error 4 mean the drive has failed? The disk activity LED is on a lot.

Thanks. ~Bob

[microsvc]

Bitor, you assisted me recently with the 500gb no longer being recognized in the 1000 and I took the recommendation of you and other posts and got a 320gb.

I can get to the web login. The drive is not visible in explorer any more, it has a big red "X". The web site on the drive is not accessible either.

Am I right to think that the drive has hit a brick wall as rpmurray suggests?

SpinRite doesnt do anything for me on these drives, I've tried it in the past. I'll give the fix fatal errors a try later and report back, but I'm not hopeful.

I have a smaller drive I can stick in there and format it if all else fails. Smaller, but it I could deal with it.

Thanks!

[bitor]

The "X" means it is off line. That tells me that the drive might have failed. Run HDD Regenerator on the drive and see if it fixes the hard drive and reinstall it in the Snap1000. HDD Regenerator is the best for fixing drives. I haven't used Spinrite in years for a HD fix.

If you have a warranty on this drive then return it to the mfg if the HDD Regenerator doesn't work. It looks like you got a bad drive.

I've got a 250GB Snap1000 that had the fsck problem at one point, but mine cleared up after I did a SnapOS upgrade.
I also have another one with a 500GB in a Snap1000 and have had no problems with it. Both use non server raid drives with OEM power supplys. But to be safe it's better to get a server drive. But I know how price can affect buying.

If you are not concerned about the data and you cannot fix it with HDD Regenerator and I hope you backed up your data then slap it in to a windows box and see if it works with an installation. Paying a data recover lots of money for most ppl is out of the question.

The only file limitation that I'm aware of is the SnapOS of 1.2TB approx. then power supply problems can go a long with bigger drives.


Let me know how it goes, but I think you got a bad drive.

bitor

Originally Posted by microsvc

Bitor, you assisted me recently with the 500gb no longer being recognized in the 1000 and I took the recommendation of you and other posts and got a 320gb.

I can get to the web login. The drive is not visible in explorer any more, it has a big red "X". The web site on the drive is not accessible either.

Am I right to think that the drive has hit a brick wall as rpmurray suggests?

SpinRite doesnt do anything for me on these drives, I've tried it in the past. I'll give the fix fatal errors a try later and report back, but I'm not hopeful.

I have a smaller drive I can stick in there and format it if all else fails. Smaller, but it I could deal with it.

Thanks!

[blue68f100]

Not all 1000's are the same, there is 3 different HW versions that I know of. If you can access the GUI with out a HD you have the later v2 or v3. There is a memory size difference between the earlier models. If you only have 32meg your size and OS is very limited. With the 64 meg version it will run the latest v4 SnapOS. If you only have 32 meg you do not want to go past v3 due to memory restraints.

Since it has reported and fatal error, did it go into PANIC mode. LED Blinking rapidly. If in Panic some time a system reset starts it backup and it will try running fsck again. Now if its trying to run fsck and you keep interupting it, you make it worst. The SnapOS will not mount the share till it knows the drive is clean with NO Errors. Now with large HD and A very very slow CPU, this can take days and even weeks if it is low on ram or have a HD >90% full.

Now remember these are not very fast so Give the SnapOS time to execute what it's doing. If you interrupt it it just make the process take longer.

If your testing with the case off, put a book over the top for the cooling to work correctly. Other wise you will over heat the cpu and HD.

[microsvc]

With the nature of this error "Cannot Read: Blk 82304 Disk 10000", I chose to replace the hdd with a 160gb I had. It works now and all my data has been restored (Yay for backups).

I tried to log on with no drive and could not, but I am using OS v4, so I dont know what version HW I'm using.

The Snap is back up now and I thank all for your help and info, till next time...

~Bob

[bitor]

microsvc,

Glad to hear you had a back up. Can you tell me what make and model the drive you had in it that did not work? Did you try to use this drive to see if another os would install? Did you check the drive for errors with any utils?
I've never heard of anyone being able to connect via the webgui without a drive & OS installed with the Snap1000 models before.

I personally like the OS...simple, free of any known virus, and very dependable. To bad they stopped developing the OS. They left some limitations to it. Which is the death of it. It would be nice if somebody made a recovery software for this os, but that will never happen for the public.

Best of luck with it. I hope it works a long time for you.


bitor

Originally Posted by microsvc

With the nature of this error "Cannot Read: Blk 82304 Disk 10000", I chose to replace the hdd with a 160gb I had. It works now and all my data has been restored (Yay for backups).

I tried to log on with no drive and could not, but I am using OS v4, so I dont know what version HW I'm using.

The Snap is back up now and I thank all for your help and info, till next time...

~Bob

[Phoenix32]

Originally Posted by bitor

I've never heard of anyone being able to connect via the webgui without a drive & OS installed with the Snap1000 models before.

Well, now you have...

Originally Posted by bitor

simple, free of any known virus, and very dependable.

No real advanced capabilities, has some old exploits, quirky, slow, very limited storage, did I say slow, no serious internal recovery abilities, really slow, a drive failure more often means bye bye data, and is about the speed of grass growing.

[microsvc]

Bitor, The drive that gave me issues is a Maxtor MaXLine II 320gB ATA/133. I chose this one because it was advertised to have 16mb of memory. Further research showed it to have 2mb memory (and runs at 5400rpm), which may explain some of the issues I had (based on your comments). I put the drive on a USB cable and Windows recognized it and formatted it to 301gb. Looks like the drive "may" be good, just didn't have the ram on board for the Snap OS. Comments?

The 160gb I installed is a 7200rpm model and has 8mb on board and formatted quickly. I'm hopeful.

~bOb

[bitor]

This is directed to microsvc and no one else.

Ok, thanks for getting back with me. I don't use Maxtor's because generally speaking, they do not have a good track record with snap servers. Not to say that some will not work. I've had the best luck with either Seagate or WesterDigital(WD) in that order. The 5400RPM is good for keeping the heat down inside the snap server. Where did you find a 5400RPM 320GB drive?
Anyway, at least the drive can be used for something else. I would test the drive first to make sure everything is ok before using it for anything. Or just sell it.
I just noticed that it's a ATA/133Mhz. That's not good. I've had no luck in a Snap1000 with most ATA/133Mhz speed hard drive. I try to only use ATA/100Mhz on my Snap1000 because of this.
Do you know what speed the 160GB drive you have in your snap1000?

Best of luck with your Snap1000. You will find it to be very reliable and cannot beat the price on them now. It's a very well made little server and dependable too.

bitor

Originally Posted by microsvc

Bitor, The drive that gave me issues is a Maxtor MaXLine II 320gB ATA/133. I chose this one because it was advertised to have 16mb of memory. Further research showed it to have 2mb memory (and runs at 5400rpm), which may explain some of the issues I had (based on your comments). I put the drive on a USB cable and Windows recognized it and formatted it to 301gb. Looks like the drive "may" be good, just didn't have the ram on board for the Snap OS. Comments?

The 160gb I installed is a 7200rpm model and has 8mb on board and formatted quickly. I'm hopeful.

~bOb

[blue68f100]

Originally Posted by microsvc

With the nature of this error "Cannot Read: Blk 82304 Disk 10000", I chose to replace the hdd with a 160gb I had. It works now and all my data has been restored (Yay for backups).

I tried to log on with no drive and could not, but I am using OS v4, so I dont know what version HW I'm using.

The Snap is back up now and I thank all for your help and info, till next time...

~Bob

Sounds like you have a late model 1000 (v2) just before the 1100's were released. This model was released in two memory configurations, 32 or 64meg. With Most being 64meg. This model uses the 2200 image file. If you do not have an image file for recovery process I would recommend making one.

[bitor]

This is directed to blue68f100 and no one else

Doesn't the Snap1000s v1 & v2 = 32MB and the Snap1100v3 = 64MB? I talked with an Adaptec tech. a while back and he said they did not make a Snap1000 with 64MBs. According to him, Snap1100 only had 64MBs.
One of mine is a Snap1000 v2 and I guess you can use either the 1000 image or 2000 image. In my case, I used the 1000 image and not the 2000 image. I've never used a 2000 image for any of my Snap1000 before.

bitor

[Phoenix32]

Originally Posted by bitor

This is directed to microsvc and no one else.

I just noticed that it's a ATA/133Mhz. That's not good. I've had no luck in a Snap1000 with most ATA/133Mhz speed hard drive. I try to only use ATA/100Mhz on my Snap1000 because of this.

BS, it just means you have configured the drives wrong with most of the ATA133 drives you have used.

Originally Posted by bitor

This is directed to blue68f100 and no one else

I talked with an Adaptec tech. a while back and he said

My experience has been that when it comes to the Snap OS units, David often knows more than the "adaptec techs" do.

[blue68f100]

Doesn't the Snap1000s v1 & v2 = 32MB and the Snap1100v3 = 64MB?

NO, in general yes, but there were some 1000 released with 64 meg just before the 1100 was released. These were actually 1100 models under the hood, they were just using up all of the old hardware up.

Your source may not have access to the early Meridian units info. Or did not work on the 1000's.

V1 & V2 had flashram for holding the OS. I have one of these at the house that was given to me. The 3rd version had a different mother board same as the 2200. The new MB version was availiable in 2 ram configuration, 32 or 64. There was a very low count of 1100 with 34meg, soon after it's release the ram was bumped up to w/64meg so JAVA could be run. You do remember JAVA don't you ???? This model required the 2200 image file. With all of these different version it was a shot in the dark to know what image file to use, most used the 2200. If you had a 1000 v1 you could access the gui without a HD. All you needed to upgrade was to install a New HD and setback while it transfered the OS to HD and format the HD. Worked just like the 2000 v1 & v2 model. The 1100 used the same MB as the 2200 with the missing header for the second HD. So if your works only on a 1100 image file you have an older MB. There were some 1100 that actually had a full/complete 2200 MB.

Remember that the SnapOS only reports models as 1x00, 2x00 and 4x00.

All of these SnapOS units have twerks. This is why I put the SnapOS FAQ's together may years ago.

btw.. This is for all who want to learn about the history of these older units. This is an open forum for all to read. So I edited your post.

[bitor]

blue68f100, good information. Thanks. That was a nice little history on them. I guess you have seen a lot of them to know all of this information. I was thinking about trying to install a 2000 image just for the heck of it. Wondering what will happen or what it will do. Did you have a chance to look at my Snap1000 mirror thread using a 1000 image? It is a 1000 mobo by the way. What do you think? I know it's out dated(snap1000), but just a fun thing for me to work out.
Extra info from my notes:
SnapOS can have no more than 32,000 files per folder and no more than 250 shares. Doubt anybody will get to that point.
Also, Mac9.22 Must move whole folders and not single files when tranfering from SnapOS.
Maybe you already knew this information.

bitor

[blue68f100]

If you install the wrong image it will fail to boot.