I've seen it discussed several times in this forum, but I thought I would bring it up again.

 

Will unRAID support dual parity anytime in the foreseeable future?  For my purposes (media server for 1-2 rooms), dual parity is the ONLY thing unRAID lacks when compared to more traditional RAID setups.  I would REALLY REALLY love to see dual parity in future versions (5.0?) please please please.

 

It took me forever to rip my DVD collection.  Dual parity could dramatically reduce the chances of having to do that chore ever again.  On a side note, I have a nice little batch file that I used to rip most of my DVDs with DVDFab and MakeMKV.  If anyone is interested let me know and I can post it over in the Lounge.

Post it up!

 

It took me forever to rip my DVD collection.  Dual parity could dramatically reduce the chances of having to do that chore ever again.  On a side note, I have a nice little batch file that I used to rip most of my DVDs with DVDFab and MakeMKV.  If anyone is interested let me know and I can post it over in the Lounge.

 

I'd be interested as I have been look for just that solution (I use MakeMKV).  I plan to install the blu-ray drive in my server case.

Post it up!

 

Sure thing.  I'll post it this evening when I get home from work.

 

 

The parity drive is no more important then any other data drive.

Until we have extra parity, I would suggest getting a cache drive the same size as your parity drive.

Then if any drive in your unRAID array fails, run the mover, stop the array.

Re-assign the cache drive to the failed drive's slot and let unraid rebuild onto the new drive until your replacement comes in.

The parity drive is no more important then any other data drive.

Until we have extra parity, I would suggest getting a cache drive the same size as your parity drive.

Then if any drive in your unRAID array fails, run the mover, stop the array.

Re-assign the cache drive to the failed drive's slot and let unraid rebuild onto the new drive until your replacement comes in.

 

Weebo ... I understand that the parity drive is no more important than the data drives.  My interest in dual parity is simply to allow a complete recovery from a two-disk failure.  Complete data recovery from a two-disk failure is not currently possible with a single parity drive.  As our disk arrays get larger, the likelihood of a two-disk failure is increasing.

 

I don't intend to use my server for any 'mission-critical' type data.  My plan is to use unRAID as purely a media server.  However, I really don't want to re-rip hundreds of DVDs ... ~200 DVDs in the event of a data+parity disk failure, ~400 DVDs in the case of a data+data disk failure.

 

 

ok ... I posted my script in the Lounge. 

Now ... how about that dual parity?  Is it in the cards?

 

 

My interest in dual parity is simply to allow a complete recovery from a two-disk failure.

 

How does this work?

 

My interest in dual parity is simply to allow a complete recovery from a two-disk failure.

 

How does this work?

 

 

There are several different algorithms that can be used to implement dual parity.  The most common, which is commonly used in RAID 6 arrays, is Reed Solomon coding.  For any fellow academic types, here is a decent short paper on implementing Reed Solomon in "RAID-like" systems.

http://www.cs.utk.edu/~plank/plank/papers/CS-96-332.pdf

 

The downside of dual parity, is that you must allocate a second parity disk.  I don't really see this as much of a downside since disks are pretty cheap these days.  I'm very willing to pay for a little more peace of mind.

 

 

Another interesting paper...  http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.88.332&rep=rep1&type=pdf

EVENODD technique for "recovery of 2 failed disks" describes read-modify-write timing and how it offers better performance over REED-SOLOMAN.

Yes please!  This is the ONLY thing stopping me from pushing the button on building my own unRAID.  I've got 13TB worth of data and growing rapidly.......

I agree that a second "parity" disk would be a nice feature to provide extra protection in the event of a true double disk failure scenario.

 

However, as far as I can remember, there has not been any documented cases where this has happened. Users HAVE lost data in other ways that are far more likely to occur. Just don't want anyone to feel that dual "parity" is the holy Grail of protection.

There's an incidence of this on the forums right now...

There's an incidence of this on the forums right now...

 

Maybe. I encouraged those affected to document dual disk failures to Tom as evidence that this is actually happening.

Agreed, it needs to be documented.

As the arrays grow in size, dual parity or hot spare become more important.

Maybe a poll for host spare or dual parity should occur.

 

Food for thought.

I think if people are so worried about multiple drive failures, they are not using quality hardware, maintenance procedures or considering diversification.

 

If your drive is so old that you are using it till it fails and worried about it's failure. then consider peace of mind and a more recent hardware investment.  

If there's an array with umpteen drives that are older then 3 years old, you may want to consider an upgrade or consolidation.

 

Maintenance procedures.  monthly parity checks are very important. It exercises your drives.

What is needed from the unRAID end is a monitoring of the drives and smart values.

Not just the temperature, but smart health. The downside of this is getting warnings off the server.

hence we need monitoring and emailing functionality.

 

Last but not least is diversification. Purchasing many drives at the same time is recipe for disaster.

One bad batch could be the case for multiple drive failure very easily. I've seen it occur multiple times.

 

Now with dual parity, you are protected with multiple drive failures. yet consider. All the drives still must be spinning until you replace the failed drive.

 

If you are doing monthly parity checks, your drives are being exercised already.

Now with a hot spare, if a drive does fail, all the drives will spin up to fill in the missing drive.

If unRAID were to start rebuilding the missing drive onto the designated hot spare, recovery and protection is actually quicker.

 

It all depends on how long you are willing to leave your array unprotected and at what cost.

 

Also consider.

An idle hot spare (that maybe is exercised during monthly parity checks) vs an active second parity drive that gets the same wear and tear as the regular parity drive.

Tom saw the need for dual parity back in 2008 and said that he would be implementing Q-Parity.

 

I don't think there is need to document multiple disk failures any more than single disk failures.  As the number of disks in an array grows, the chance of multiple disk failures increase just like the need for dual parity.

 

As for hot spare; it would be nice, but in addition to dual parity -- not as an alternative.  In Raid 5, the process of rebuilding an array is generally where I've had my second disk failures.

 

Also consider.

An idle hot spare (that maybe is exercised during monthly parity checks) vs an active second parity drive that gets the same wear and tear as the regular parity drive.

 

Now that is an interesting idea...got me thinking.  I didn't want to take this thread too far off topic, so I started another one:

 

Warm Spare disk assignment slot

 

 

 

I agree that a second "parity" disk would be a nice feature to provide extra protection in the event of a true double disk failure scenario.

 

However, as far as I can remember, there has not been any documented cases where this has happened. Users HAVE lost data in other ways that are far more likely to occur. Just don't want anyone to feel that dual "parity" is the holy Grail of protection.

 

Nope ... dual parity certainly isn't the holy grail, but it does offer a little bit more peace of mind.  If I can spend $100 to feel a little bit more comfortable with the security of my data ... I'm all over it. 

 

 

Maybe a poll for host spare or dual parity should occur.

 

Why choose between dual parity and a hot spare?

Why can't we have both?

 

Keeping my data safe is very important to me.  I want as many layers of security for my data as possible.  But, I don't really want to maintain two separate RAID systems.

 

 

Tom saw the need for dual parity back in 2008 and said that he would be implementing Q-Parity.

 

I don't think there is need to document multiple disk failures any more than single disk failures.  As the number of disks in an array grows, the chance of multiple disk failures increase just like the need for dual parity.

 

As for hot spare; it would be nice, but in addition to dual parity -- not as an alternative.  In Raid 5, the process of rebuilding an array is generally where I've had my second disk failures.

 

 

The squeaky wheel gets oiled.  If there are enough cases documented where a rebuild fails due to a second drive failure I'm sure it would provide leverage in the development effort.

Maybe a poll for hot spare or dual parity should occur.

 

Why choose between dual parity and a hot spare?

Why can't we have both?

 

Keeping my data safe is very important to me.  I want as many layers of security for my data as possible.  But, I don't really want to maintain two separate RAID systems.

 

1. Cost - 2 extra drives per array. (for people who are budget and space minded)

2. Programming time and effort (While I believe both provide value, the time and effort involved needs to be prioritized. if this were to happen in stages, Hot spare, then q-parity or vica versa, there are all sorts of efforts and betas that need to occur).

 

Also, I never said one without the other.

Keeping my data safe is very important to me.  I want as many layers of security for my data as possible.  But, I don't really want to maintain two separate RAID systems.

 

I hope people are not using multiple parity as a substitute for backup. It isn't.

The squeaky wheel gets oiled.  If there are enough cases documented where a rebuild fails due to a second drive failure I'm sure it would provide leverage in the development effort.

 

I was kind of bit by this just recently.  My Parity drive went down with some write errors.  I reseated the cables and it came back but it wanted to do a parity check so i let it.  During that time one of my data drives went down.  So now the server thinks parity is bad and I have a data drive that went down. Hopefully i can bring the drive back up long enough to pull the info off of it...

 

 

I hope people are not using multiple parity as a substitute for backup. It isn't.

 

Everyone knows that though don't they?  But it's certainly not cheap to have a true backup for 15TB+ worth of data, and unless it's offsite you have still have issues with fire, flood or theft.

I agree that a second "parity" disk would be a nice feature to provide extra protection in the event of a true double disk failure scenario.

 

However, as far as I can remember, there has not been any documented cases where this has happened. Users HAVE lost data in other ways that are far more likely to occur. Just don't want anyone to feel that dual "parity" is the holy Grail of protection.

 

I'm a SAN admin for an enterprise.  It happens quite frequently.  personally, I've had a double drive failure a few times.  We build the arrays where we can to minimize this.  However, once in the 7 years we've been running the SAN at my office, we've had a double drive failure on a single array.  Thank GOD it was a raid6 array, or we'd have lost data.  We have had 5 drives fail at once in a raid5 storage controller, but they failed in a way that there were only individual drives in the 5 arrays where a drive failed.

 

From talking to my peers, and reading forums, it happens quite frequently.  Even in the enterprise arena, where drives are more reliable than the consumer drives that would likely be in an unraid array.

 

So for my two cents, a second parity would be worth every penny of a second drive.  It's "only movies/mp3/non-critical data', but my time alone to do it would be worth it.

 

I've got a study somewhere from one of our san vendors that has the statistics of a drive failure occurring that will cause data lose in different raid types... It was rather interesting, iirc, the double parity failure was something like 328 years.  We had the double drive failure after that presentation.  it was interesting to talk to the people who gave the presentation and say "should I go play the lotto?"