New UNRAID user here but I've been lurking these forums for quite some time before finally taking the plunge this past week. That said, I have what I think is a drive compatibility issue but am hoping someone can help diagnose the exact nature of the problem so that I do not repeat it again in the future.

 

Quick background info:

I'm running UNRAID 5.0Beta7

All 6 hard drives were brand new, never used prior to my installation and all 6 passed the Diagnostic Tests of their respective manufacturers prior to install. (Drives consist of 3 Seagates, 1 Western Digital, 1 Hitachi, and 1 Samsung)

The drive in question is the Samsung (model# HD155UI, 1.5TB) .

 

OK...

 

1)My first few attempts to allocate and format the drives led to an unresponsive server that required rebooting. After a few reboots I noticed during either the format or parity sync phase that one of the drives (Samsung HD155UI) had an unusually high number of errors (>7,000) as well as a red dot next to it.

 

2)Per the UNRAID manual the red dot indicated a problem with the drive and that it was "removed" from the array. At this point I pulled the drive, reset the array with the other 5 drives, and it has been smooth sailing ever since.

 

3)Concerned that something may have happened to the drive since I had first put it through the diagnostic test a few weeks ago I reran Samsung's test and it passed with flying colors.

 

4)After searching unsuccessfully these forums to see if anyone else had had trouble with this drive I convinced myself it might of just been a "first time configuration fluke." Therefore, I tried adding the drive as per the instructions (stop the array, power down, add drive, power on, assign drive, start array...) At this point the server went unresponsive again, I lost web access, and lost access to the shares. I did have telnet access and could see the Flash drive on the network but that's about it.

 

5)After a couple of reboots the server finally came alive again with the Samsung drive "unassigned."

 

So, can I just assume this drive is incompatible with UNRAID? Is there something I am doing wrong preventing me from adding this drive to my array (e.g. firmware update). Or, is it a possible defect in the drive that the diagnostic test did not pick up? And lastly, is this the normal behavior of UNRAID if it encounters an incompatible drive?

I've never seen that behavior with unRAID and hard drives, including 2TB versions of the F4. How does the drive behave on other systems? Other drives in the same position on this unRAID?

 

Don't put too much faith in manufacturer confidence tests.

Not a compatibility issue. Did you preclear the disk?  If not, unRaid will clear it when you add it to your array. Will take a number of hours to complete during which the array will be non-responsive.

I had not pre cleared the drive. However, I hadn't pre cleared any of the new drives prior to my first time setup yet, this is the only drive that caused problems.

 

I did attempt to pre clear the drive yesterday but the server went unresponsive and needed a reboot after only about 20 minutes. That said, I initiated the pre clear process while the array was busy copying files from my older server. Therefore, I just figured I needed to retry when the array was "less busy." Is this expected behavior?

 

My system specs are:

BioStar A760G M+

1GB Kingston Value RAM

Sempron 140

Preclearing / clearing is not needed when defining a new array - it is just needed to add a disk to an existing array.

 

No, this is not expected behavior.

 

You should run a smart report on the Samsungs drive. Post a syslog. Often these problems trace back to a bad or loose cable.

I attached the system log generated shortly after I lost web access after attempting to start the array last night with the Samsung drive.

syslog.txt

System Log was attached in my last post. Hopefully, a knowledgeable person could take a look and see what the issue with the drive was/is. I am nearing the end of the 30 day return/exchange period and am trying to figure out what the issue is to determine next steps.

 

I have two of these Samsung Drives (the other was DOA and I'm awaiting a replacement) but don't wan't to risk not being able to exchange them if it is a compatibility issue. Any help would be appreciated.

It sees it as a new disk:

 

Jun 26 21:38:37 Tower kernel: md: import disk4: [8,80] (sdf) SAMSUNG_HD155UI_S2HEJ1DB303167 size: 1465138552

Jun 26 21:38:37 Tower kernel: md: disk4 new disk

 

And is clearing it:

Jun 26 21:38:38 Tower emhttp: clearing disk4...

 

This process can take hours, during which time UNRaid shares aren't available.  If you want to avoid the extended downtime when adding a new disk, then use Joe's preclear script to clear the drive outside of the array.  Then when you add it in, all UNRaid has to do is format it, which should take less than a minute.

 

As to why you didn't see this behavior when you added the previous drives, I can't say.  But best as I can tell from your syslog, everything is working normally.

I know this isn't helpful, but why do new users always seem to jump in with a 5.0 Beta?

 

I mean, you are jumping into something totally new, where you have no experience.  You are jumping into a Beta platform that has bugs and unknown issues (along with the known issues)  that you potentially have to deal with.  Then you don't know enough about the system to try to workaround or fix the bugs.

 

Why don't new users start with the most recent stable version, and least for the first few months?  I never understood...

I know this isn't helpful...

 

At least you know.

 

I mean, you are jumping into something totally new, where you have no experience.  You are jumping into a Beta platform that has bugs and unknown issues (along with the known issues)  that you potentially have to deal with.  Then you don't know enough about the system to try to workaround or fix the bugs.

 

Why don't new users start with the most recent stable version, and least for the first few months?  I never understood...

 

I can't speak for others but I opted for the Beta because it is not for my primary server. It is a project I'm working on to one day be my primary server (a Beta in its own right). I figured by the time I'm ready to make it my primary server, 5.0 will be a stable release. And, while UNRAID might be different, often times major releases are a pain to upgrade. Therefore, I opted for a likely cleaner upgrade path down the road. That said, I am keenly aware of the risk associated with Beta software. However, the drive issue did not have the characteristics of an issue that could be dismissed as simply a Beta related bug. So, I came here and asked for help.

 

Why some users view newbie help requests as an opportunity to complain, I'll never understand

I've had the opportunity to try and preclear a few other drives and the one common denominator I have found is the SASLP-MV8, which appears to have caused some other people grief as well. And, to my discredit I forgot to include in my initial spec list.

 

Since i have not yet installed all the drives which will utilize the on board sata ports I am perfectly content using those to preclear the disks. Topic closed.