Are unraid users more careful?

[gabbott]

So I've been reading about unraid for well over a year and finally made the plunge. One thing that kept me back from trying for a while was it seemed like one difference between unraid and other software NAS solutions was it was more picky about its hardware. Also there seemed like a focus in the forums and wiki that you must run things like preclear and parity checks over and over to verify things are in order. It just seemed like an emphasis on ongoing verification of one's data.

 

I've never noticed such an emphasis on this from any other online community or in documentation of other solutions. So my perception started to be that you did all that because there wasn't a complete trust in unraid as it if was somehow different than say freenas, openfiler, WHS....

 

So that perception also made me a little leery to give it a try. But now that I think about it, one really shouldn't "trust" any file server solution completely and maybe this focus on data integrity isn't something that has to be done inherently of using unraid, that is where my perception was off.

 

So just thought I'd throw this out there for discussion...

 

[Joe L.]

I have trust in the unRAID software.  I do NOT have any trust in ANY hardware that has not been tested in a specific person's server, even if the hardware is on the compatibility list.

 

UnRAID really does not care too much about the specific hardware... It will run on almost any motherboard.  To keep it reasonably small in size, so it will still fit in a small USB flash drive, it does not have every driver for every possible network card ever made.  It does contain most of the common ones for Gigabit LAN chipsets.    (Some boards with brand new chipsets will not have a driver... the solution for some is to then use a add-on network interface card.)  The same goes for SATA controller chipsets, most common ones are supported.  Some are not.  For that reason, the hardware compatibility list exists.

 

In helping people with random parity errors over the past year or so, the errors have all been hardware related.

We've seen bad motherboard chipsets, bad CPU, bad RAM, bad voltage and timing settings for RAM, even a bad hard-disk.

 

It is because unRAID owners piece together servers out of parts they have laying around is why it is highly recommended to test their server for reliability before committing their precious data to it.  99% of unRAID users will not run into any problems with their old hardware, but the very fact that it is old hardware might lend it to have problems from old electrolytic capacitors and poor voltage regulation.  Even new hardware can have problems... so it is just as important to test a new motherboard with the specific RAM you purchased, with the timings, voltage, and clock speed you've set in your BIOS.

 

I personally hate to learn a person has loaded their entire media collection onto the unRAID server, only to learn it is corrupt, or that subsequent parity checks return errors, because they did not perform any checks of the hardware being used.  Unless the array is powered down without first stopping it, you should never see a parity error ... If you do, you have a hardware issue, somewhere and it needs to be addressed.

 

Yes, we are more careful (or at least some of us are)   Should the other NAS be as careful???  if they care about the data they should.

[gabbott]

 

Yes, we are more careful (or at least some of us are)    Should the other NAS be as careful???  if they care about the data they should.

 

And that was the point I realized. Not to say users of other NAS software aren't, but here there is an emphasis on that here, which I think is great.

[prostuff1]

So I've been reading about unraid for well over a year and finally made the plunge. One thing that kept me back from trying for a while was it seemed like one difference between unraid and other software NAS solutions was it was more picky about its hardware. Also there seemed like a focus in the forums and wiki that you must run things like preclear and parity checks over and over to verify things are in order. It just seemed like an emphasis on ongoing verification of one's data.

 

I've never noticed such an emphasis on this from any other online community or in documentation of other solutions. So my perception started to be that you did all that because there wasn't a complete trust in unraid as it if was somehow different than say freenas, openfiler, WHS....

 

So that perception also made me a little leery to give it a try. But now that I think about it, one really shouldn't "trust" any file server solution completely and maybe this focus on data integrity isn't something that has to be done inherently of using unraid, that is where my perception was off.

 

So just thought I'd throw this out there for discussion...

 

 

I like to think of it as diligent

 

I really never paid attention to drives before reading into unRAID.  I have had to send multiple drives back (all Seagate) for multiple problems.

[Chris Pollard]

I wasn't any less careful back when my main storage was FreeNAS.  Daily smart checking,  weekly md5 checks.

 

I'm sure the forums skew perception of the product anyway,  I'm guessing the majority of users just set it up and forget about it.  The only reason for them to come here is if they have a problem.

[Rajahal]

Becoming an unRAID user has certainly lead me to be a bit more careful with my data, and to give more thought to practices like burning in hardware before putting it to use.  All of my most important data is backed up on my unRAID server, as well as on every one of my and my family's computers using Crashplan. 

 

However, I'm still not as careful as some users here.  I do routine maintenance, like checking the web management page daily or semi-weekly (looking for red balls/failed drives) and monthly parity checks, but that's about the extent of it.  I don't monitor the SMART status of the drives in my unRAID server, and even if I did I wouldn't preemptively replace dying drives.  I trust in the unRAID software to save my data when a drive dies, so I would rather just let it die on its own and deal with replacing it then as opposed to constantly monitoring my drives and trying to decide myself when to replace them.  Past experience tells me that even a drive which SMART tells you should be replaced can last for months or years past that point.  My experience with unRAID also tells me that my data collection grows at such a rate that I usually end up replacing a drive with a larger drive out of need for increased capacity well before the smaller drive starts to wear out.

 

I allow myself to be a bit more lax in this regard because there isn't any data on my unRAID server that isn't replaceable.  Sure, I would rather avoid losing data, but even if I did I would be able to get it back again with enough time and effort.  So in my situation, unRAID is more of a convenience than a fail-safe.

[WeeboTech]

I trust in the unRAID software to save my data when a drive dies, so I would rather just let it die on its own and deal with replacing it then as opposed to constantly monitoring my drives and trying to decide myself when to replace them.  Past experience tells me that even a drive which SMART tells you should be replaced can last for months or years past that point.

 

It may be the case, that a SMART failure replacement warning is pre-emptive, but it's also a condition that manufactures will replace a drive within warranty. If I have a chance to replace a drive in warranty vs waiting for a failure and having to purchase a new one, might as well get a refurbished drive for the time being.

 

The only downside these days is that Seagate charges for advanced RMA. (This is something new I just learned about.. which I think is bogus).

 

In the days of Maxtor, as soon as I received a non healthy SMART status. I would run the maxtor tools, grab the error code, get the advanced RMA on it's way and have a replacement before the drive died or at least on it's way.

 

There really is no reason to delay once you have a SMART health error and you are within warranty.

Chances are you could forget about it, and miss out on the warranty replacement (I've done that a couple times too.. hahahaha).

[abeta]

The only downside these days is that Seagate charges for advanced RMA. (This is something new I just learned about.. which I think is bogus).

 

Yeah, same here. I would think that if I'm willing to give them my credit card as collateral they should advance ship it but they won't. They said the charge is because they're paying for 2 day shipping which depending on your view may or may not be worth it.

 

In my case $20 is around 20% of a new drive which makes it bogus to me.

 

I also think its a fairly recently policy too as I've swapped out Seagate drives early last year and don't recall any fees.

[prostuff1]

The only downside these days is that Seagate charges for advanced RMA. (This is something new I just learned about.. which I think is bogus).

 

Yeah, same here. I would think that if I'm willing to give them my credit card as collateral they should advance ship it but they won't. They said the charge is because they're paying for 2 day shipping which depending on your view may or may not be worth it.

 

In my case $20 is around 20% of a new drive which makes it bogus to me.

 

I also think its a fairly recently policy too as I've swapped out Seagate drives early last year and don't recall any fees.

 

I think the charge for a drive that is under warranty is also bogus.  I had to pay for it once on one my seagate drive.  I got the replacement that failed 1 month later and got a free one to replace it; I was just within the month time frame to not have to pay for the advanced replacement.  I have a 1.5TB drive that keeps resetting (hard resetting link errors) when I use it in unRAID and I can not for the life of me get it to stay stable to get a report from it.  I have a few other tricks I am going to try on it but for now I have disconnected it and removed it from the array.  As soon as the data cable was unplugged and unRAID was rebooted I have not had any troubles with the system.

[WeeboTech]

In my case $20 is around 20% of a new drive which makes it bogus to me.

 

I also think its a fairly recently policy too as I've swapped out Seagate drives early last year and don't recall any fees.

 

It might be because they have received so many returns.

For those with firmware issues, maybe they are just sending back the same drive reflashed.

I dunno.. But I have a seagate that has failed and now I have to go through the process of ship out and wait for a replacement *sigh*.  It does sort of make me want to just upgrade the drive.

 

In any case, are unRAID users more careful? I dunno, I've been using linux software raid since the days you had to run a command before there was hot online resync/check/reconstructions.

 

It was painful if you shutdown uncleanly. You would have to wait for the whole array to resync/repair before your server came up.

 

When drive space exceeded practical magnetic tape limits, I knew I had to do something different.

When drive space exceeded practical RAID1 limits, I had to do something different (unRAID). ;-)

[SSD]

RAID solutions stripe data across all the disks. If something is wrong - like a disk failure - you will know quickly. unRaid is different. Drives could fail and parity could be completely wrong and accesses from good drives would work at full speed without problem.

[abeta]

 

I think the charge for a drive that is under warranty is also bogus.  I had to pay for it once on one my seagate drive.  I got the replacement that failed 1 month later and got a free one to replace it; I was just within the month time frame to not have to pay for the advanced replacement.  I have a 1.5TB drive that keeps resetting (hard resetting link errors) when I use it in unRAID and I can not for the life of me get it to stay stable to get a report from it.  I have a few other tricks I am going to try on it but for now I have disconnected it and removed it from the array.  As soon as the data cable was unplugged and unRAID was rebooted I have not had any troubles with the system.

Not to derail this thread. I don't know if this is the solution but check your firmware. I was having a similar problem with SD1A, I upgraded to SD1B and I haven't seen the hard resetting link errors yet. It's only been a few days.

 

As far as being more careful goes...I'd say yes and no. Yes in the sense that I've never really had a centralized storage system for data and backup at home and since I spent a bit of time getting the files over, etc. I do check the status page regularly (at least daily) and looking at the logs for obvious errors.

 

No, in the sense that I don't think it'd be different regardless of what system I'm using. I'd probably keep checking it as it is new to me (less than 3 months) until I get that warm and fuzzy feeling of trust then I just won't worry about it.