Category Archives: RAID

Fundamentals of Storage Systems – RAID and Hard Disk Reliability, Under the Covers

In the last RAID article we covered the basics. This is a little deeper dive into the underlying mechanics of RAID. Exactly what it does, how it does it and what it doesn’t do that people assume it does. I sited David Patterson, Garth Gibson, and Randy Kats and their work at UC Berkley on RAID. They show something I’ve talked about before the “Pending I/O Crises”. Of course it isn’t pending anymore, its here. One of the concerns has to do with Amdah’s Law and speeding up execution with parallel operations. As processors and memory speed up hard disks are still an order of magnitude slower. Another aspect is Kryder’s Law, which like Moore’s Law, is a estimation of capacity growth of hard disks over time. Kryder’s Law is starting to slow down just as Moore’s law is. The problem with hard drives has never really been capacity, its speed. As areal density increases you do get an increase in data throughput, there is simply more data per square inch on the disk. You also get an improvement in I/O’s, tracks are closer together. We haven’t broken past the 15k barrier yet. I’ve still got Seagate Cheetah 15k.3 drive from 2002. It has a max sequential throughput around 80 MB/sec. I doubt we will see spinning disks faster than 15k. This is a real problem for scaling I/O up. Enter RAID. It’s simple get a bunch of disks and then stripe data across them. One little problem creeps up. Reliability goes down for each drive you add to the array. Using RAID 0 pretty much guarantees you will have an array failure. To overcome this We start adding some way to make the data more redundant.

Hard Disk Reliability

People make a lot of assumptions about hard drives and their reliability. Hard disks break down into two classes consumer grade, the drive you have in your desktop and enterprise, the kind usually in your servers. There are misconceptions around both. Recently, Google and others have written papers based on long term large batch sample failure rates and found the enterprise class drives don’t last any longer than consumer class. This study is perfectly valid from a physical reliability point of view. Most drives are manufactured the same way in the same plants. Not like the poor misunderstood lemming, hard disks do all jump off a cliff together. Studies have shown that there is a strong corollary to disk failure and a shared manufacturing batch. Simply put, if they are made around the same time if one has a failure there is a likelihood, around 30%, other drives in that batch will also suffer failures. So, what are we paying for with an enterprise drive besides speed? Data reliability. Enterprise level drives have more robust error correction than their consumer counterparts. On a normal hard drive the smallest piece of data that can be written is 512 bytes. This is the size of a sector. Enterprise drives usually have 520 byte sector 8 bytes are used to verify the data in that sector, this is the Data Integrity Field. DIF isn’t 100% ether. It is more reliable than a consumer drive without it. You can still have write corruption for several reasons. Misdirected writes occur when data is written to the wrong location on disk and reported as a successful write. When the system goes to access again you get a read fault. Torn pages, which we are familiar with, is when an 8k page write is requested but only part of the 8k is actually reported. Corruption outside the drive where the controller makes a bad request to write but it is a perfectly legitimate I/O request at the hard drive level. With larger drives the odds of hitting one of these errors becomes a real possibility. Enterprise drives add this extra layer of protection. Your RAID HBA may also have additional error correction. The last thing I would like to touch on is write catching. Without a battery backup, or if the cache non-volatile in nature, you will loose data on a power failure if a write is in progress.

RAID Host Bus Adapter Reliability

The adapter is as reliable as any other component in your system. Normally, the cache on the controller is ECC based. Also, you usually have the option of a battery module to supply the cache with power incase of an outage so the data in cache can be written to the array when everything comes back up. Most of the issues I have seen with RAID HBAs is almost always driver or firmware related. You may also see inconsistent performance due to write catching and the battery backup unit. The unit has to be taken off line and conditioned to keep it in top condition. The side effect is a temporary disabling of the write cache on the controller. You can override this setting on some controllers but it is dangerous proposition. I personal anecdote from my days at a large computer manufacturer, we started getting a larger volume of failed drive calls into support. We started doing failure analysis. It all pointed back to a particular batch of hard drives. That was when the drive manufacturer made a change in its drives removing very small component. It shaved a few cents off the cost but had a dramatic effect. All the drives were technically good and would pass validation. Under a enough load and attached to a particular RAID HBA they would randomly fall off line. It came down to the little component. It provided a little bit of electrical noise suppression on the SCSI bus. Some cards were effected and others chugged along just fine. This is also confirmed by the Google paper, they observed the same behavior. They also point out that 20% to 30% of all returned drives have no detectible problems. The point is validate your entire I/O stack. Any single component may be within specification but may not play well with others.

RAID Parity, Mirroring, and Recoverability

Not to belabor the point, RAID isn’t bullet proof. People rap RAID round themselves like Superman’s cape. There are several issues that all the RAID schemes in the world don’t protect against. With current hard disks in the two terabyte range it is possible to build even a small RAID 5 array and have potential for complete failure. The problem is the amount of data that has to be read for the rebuild process. Having a hot spare available reduces the time to replace a failed drive to zero but that is only part of the equation. The much larger part is rebuild time. Lets say you have a 14 drive RAID 5 array with the new two terabyte drives installed and suffer a failure. If you have no activity on the array and all the IO is detected to the rebuilt it could still take two or three days to rebuild the array. During that time you are effectively running on a RAID 0 array that is now under load. Your chance of total array failure is near 100%. RAID by its very nature assumes a failure is a hard failure. A drive goes off line and the redundant part of the system takes over. It also makes the assumption that if a write succeeds then, barring a hardware failure, the read will also be valid. Data is only validated on writes not on reads. If it was RAID 5 would be twice as slow on reads and four times as slow on writes as a single drive or RAID 0. With all the potential hidden write failures it is completely possible to have hidden corruption and not know it until it is way to late. RAID levels with striped parity are most susceptible to this kind of silent creeping corruption. It is possible that the corrupted data is in the parity stripe making it completely unusable for data reconstruction. If that particular piece of data doesn’t change you can go a very long time with a RAID 5 array with polluted parity. You know how to recover from a polluted parity stripe? Simple, copy all the data off the array, figure out which files are now corrupt and restore them. RAID 6 with its dual stripes makes it more likely to recover your data from a single parity stripe becoming corrupt. You do pay a price in write speed for that extra level of protection. RAID 1 and RAID 10 aren’t perfect ether. On a mirrored pair if the write is assumed good there is no way to validate that on read. Without a third piece of information, like a checksum, it would be a coin toss. If the read is successful there is no way to tell which drive has the bad data. It is possible to have a mirrored pair run just fine with one giving you corrupted data on reads all day long. It would manifest itself as file corruption or some other anomaly that could be difficult to track down. We are back to relying on the disk to tell us all is well. We often recommend RAID 10 over everything else for speed and reliability, and I still hold to that. RAID 10 can still suffer from a catastrophic failure due to a single mirrored pair failing at the same time. With the probability of correlated disk failures it can’t be ignored.

What Can We Do?

There are a few tools available to us that can help predict the failure of a drive or that something is wrong with the array. All modern drives support the SMART protocol. Even though Google found it wasn’t as useful and wasn’t 100% reliable, closer to 30%, some warning is better than none in my opinion. All modern RAID HBA’s also come with tools to detect parity errors. You do take a hit when you run these internal consistency checks. Just like you run maintenance on your databases via DBCC your RAID arrays need checkups too. They are a necessary evil if you don’t want any surprises one day when you have a failed drive in your RAID 5 array and can’t rebuild it. If you have intermittent problems with a drive, don’t mess around, replace it. The HBA almost always has the ability to send SNMP messages to something like nagios or HP Openview, Use it. If you aren’t running something like that usually you can configure email alerts on error to go out. Proactive is the name of the game.

Don’t take my word for it….

Short list of papers to get you started on your path to paranoia.

Google Disk Failure analysis

Original RAID Paper

NetApp disk failure analysis

CERN data corruption tests

Silent Data Corruption in SATA arrays

Series To Date:

Fundamentals of Storage Systems – Stripe Size, Block Size, and IO Patterns

If you have been following this series we have covered system buses, hard disks, host bus adapters and RAID. Along the way we also covered how to capture your IO patterns and the SQLIO tool. Now we will pull it all together.We move up the stack even further to the actual layout of the RAID stripe and the file system. How the stripe and file system are laid out on your disks has a huge impact on performance. One of the things that has really gotten some traction over the last few years is sector alignment. This one thing, if not done, could cost you 30% to 40% of your IO potential. Jimmy May has covered sector alignment in depth So I won’t hash it here again. Kendal Van Dyke also has a good series that covers offset, stripe size, and allocation units with different raid levels.

It Don’t Add Up…

Something I’ve seen, and been guilty of, is taking a drives base specifications and just multiplying out. Say the manufacturer says the drive will to 79MB/Sec minimum throughput, we have 10 drives so that is 790MB/Sec of throughput! We all know from experience that this isn’t so. What eats us up is how much slower it really can be. As we have seen throughout this series there is overhead associated to everything. Before we just throw a bunch of disks in an enclosure and press it into service it would be nice to have an idea of what the performance should be. It’s also recommended to do some of this work before you actually buy anything so you don’t have to go back to your boss and beg for more money and explain to him that your wild guess was wrong.

Always add a pinch of salt to whatever the disk manufacturer puts in the specifications. Most of the time they will be close enough. The problem lies in the fact they don’t always disclose the methods for archiving those numbers. For instance, when they report minimum and maximum throughput they are usually talking about a scan of the entire disk including all meta data stored between tracks, the best possible throughput possible. You won’t see those results in every day life. They also give you numbers that can be completely irrelevant like single sector read rates. very rarely do you read a single sector at a time. Personally, I would love if the drive makers gave the engineering specifications. I know that won’t happen, it would make my life easier though. The disk characteristics that are important are, sector size,spindle speed, seek times read and write, sequential times read and write. To a lesser extent sequential throughput in megabytes per second. With the single disk numbers we can move on to the RAID configuration.

Configuring your RAID Array

There are several factors that impact the RAID arrays ability to perform. The RAID level, size of the IO request, and stripe size. RAID level is the easy one, what kind of hits do you take on writes vs. capacity of the array. On the stripe size there is a direct corollary with the size of the IO request. If the IO request is bigger than the stripe size it will have to seek across another disk to satisfy the data request. If the IO request size is very small and random you may loose some IO performance if the requests pile up on one disk causing a hot spot. There are established calculations that you can perform to get an idea of how to configure you array. I’ve built a web page that you can use to do all the basic calculations, Disk Drive RAID Configuration Tool. These equations are base line estimates so you aren’t working completely in the dark. You can enter your own drive statistics or pick from one of 1100 hard drives in the database. This web calculator is based off of Peter Chen’s equations for estimating RAID performance and best stripe size. I’ll add more to it as I get time.

SQL Server IO Patterns and Array Performance

SQL Server works with two specific IO request size 8K and 64K in general. If you did your due diligence earlier you could also add any other request size that you saw come through. Focusing on the page size and extent size is a good place to start. Using the raid calculator tool I chose a Seagate Savvio 15K.2 drive as my base. One of the things my calculator can’t take into consideration is your system and RAID HBA. This is where testing is essential. You will find there are anomalies in every card, physical limits on throughput and IO’s. Since my RAID card won’t do a stripe bigger than 256k that is my cap for size. Reading through several IO white papers on SQL Server the general recommendation is for 2000/2005 a 64k or 128k stripe size and for SQL Server 2008 a 256k stripe size. I’ve found as general guidance, this is a good place to start as well. The calculator tells me for a RAID 10 array with 24 drives at a 256k stripe size and 8k IO request I should get 9825 IOs/Sec and 76.75 MB/Sec on average, across reads, writes, sequential and random IO requests. That’s right, 76 MB/Sec throughput for 24 drives rated at 122 MB/sec minimum. That is 2.5 MB/Sec per drive. The same array at a 64k IO request size yields 8102 IOs/Sec and 506 MB/Sec. A huge difference in throughput just based on the IO request size. Still, not anywhere near 122 MB/Sec. As an estimate, I find that these numbers are “good enough” to start sizing my arrays. If I needed to figure out how big the array needs to be to support say 150 MB/sec throughput or 10000 IOs/Sec you can do that with the calculator as well. Armed with our estimates it’s time to actually test our new RAID arrays. I use SQLIO to do synthetic benchmarking before running any actual data loads.

After doing a round of testing I found that in some cases the numbers were a little high or a little low. Other factors that are hard to calculate are cache hit ratios. Enterprise RAID HBA’s usually disable the write cache on the local disk controller and just use their own batter backed cache for all write operations. This is safer but with more and more disks on a single controller the amount of cache per disk can get pretty low. The HBA will also want you to split that between read and write operations. On my HP RAID HBA’s the default is 25% read and 75% write. In an older study I found on disk caches and cache size saw diminishing returns above 2 MB gaining between 1 and 2 percent additional cache hits per megabyte of cache. I expect that to flatten out even more as the caches get larger, you simply can’t get 100% cache ratios that would mean the whole drive fit in the ram cache or your IO request are the same over and over. Generally if that is the case you will find SQL Server won’t have to go to disk it will have what it needs in the buffer pool for reads. I find that if you have less than 20 percent write activity leaving the defaults is fine. If I do have a write heavy load I will set the cache to 100% writes.

The Results

Having completed my benchmarking I found that 128k or 256k stripe size was fine on average. Just realize that if you optimize for one IO pattern the others will suffer. Latency is also important and I have included it here as well. You find that the larger the IO request and the smaller the stripe size latency gets worse. Here are the results from my tests on a DL380 G5 with a P411 and 24 drives in a MSA 70 enclosure. I’ve included tests for an 8k to 256k stripe sizes.

As a footnote I’d like to thank Joe Handley, Ben Poliakoff, David Gosslin and Dale Davis for helping me get the Disk Drive RAID Configuration Tool together. I’m not a web guy!

WARNING! Lots of charts below!

Read 8K IO Request 24 73GB 15K Drives RAID 10 64K File System Cluster Size 1 Outstanding IO’s 8 Threads
Random	Sequential

Write 8K IO Request 24 73GB 15K Drives RAID 10 64K File System Cluster Size 1 Outstanding IO’s 8 Threads
Random	Sequential

Read 64K IO Request 24 73GB 15K Drives RAID 10 64K File System Cluster Size 1 Outstanding IO’s 8 Threads
Random	Sequential

Write 64K IO Request 24 73GB 15K Drives RAID 10 64K File System Cluster Size 1 Outstanding IO’s 8 Threads
Random	Sequential

Series To Date:

Fundamentals of Storage Systems – RAID, An Introduction

In previous articles, we have covered the system bus, host bus adapters, and disk drives. Now we will move up the food chain at take a look at getting several disks to operate as one.

In 1988 David A. Patterson, Garth Gibson, and Randy H. Katz authored a seminal paper, A Case for Redundant Arrays of Inexpensive Disks (RAID). The main concept was to use off the shelf commodity hardware to provide better performance and reliability and a much lower price point than the current generation of storage. Even in 1988, we already knew that CPUs and memory were outpacing disk drives. To try to solve these issues Dr. Patterson and his team laid out the fundamentals of our modern RAID structures almost completely RAID levels 1 through 5 all directly come from this paper. There have been improvements in the error checking but the principals are the same. In 1993, Dr. Patterson along with his team released a paper covering RAID 6.

RAID Level	Disk Required	Usable Disks	Diagram
RAID 0	2	N
RAID 0 is striping without parity. Technically, not a Redundant array of disks just an array of disks but lumped in since it uses some of the same technical aspects. Other hybrid raid solutions utilize RAID 0 to join other RAID arrays together. Each disk in the array holds data and no parity information. Without having to calculate parity, there are no penalties on reads or writes. This is the fastest of all the RAID configurations. It is also the most dangerous. One drive failure means you lose all your data. I don’t recommend using RAID 0 unless you are 100% sure losing all your data is completely OK.
RAID 1	2	N/2
RAID 1 is mirroring two disks. RAID 1 writes and reads to both disks simultaneously. You can lose one disk and still operate. Some controllers allow you to read data from both disks; others return only data from the disk that delivers it first. Since there are no parity calculations, it is generally the easiest RAID level to implement. Duplexing is another form of RAID 1 where each disk has its own controller.
RAID 5	3	N-1
RAID 5 is a striped array with distributed parity. This is similar to RAID 0 in that all data is striped across all available disks. Where it differs is one stripe holds parity information. If a drive fails, the data contained on that drive is recreated on the fly using the parity data from the other drives. More than one disk failure equals total data loss. The more drives you have in a RAID 5 array the greater the risk of having a second disk failure during the rebuild process from the first disk failure. The general recommendation at this time is 8 drives or less. In general, the larger the drive the fewer of them you should have in a RAID 5 configuration due to the rebuild time and the likely hood of a second drive failure.
RAID 6	4	N-2
RAID 6 is a striped array with dual distributed parity. Like RAID 5 it is a distributed block system with two parity stripes instead of one. This allows you to sustain a loss of two drives dramatically reducing the risk of a total stripe failure during a rebuild operation. Also known as, P+Q redundancy using Reed-Solomon isn’t practical to implement in software due to the math intensive calculations that have to take place to write parity data to two different stripes. The current recommendation is to use 8 drives or more.
RAID 10	4	N/2
RAID 10 is a hybrid or nested striping scheme combining RAID 1 mirrors with a RAID 0 stripe. This is for high performing and fault tolerant systems. Like RAID 1, you lose half your available space. You could lose N/2 drives and still have a functioning array. Duplexing each mirror between two drive chassis is common. You could lose a drive chassis and still function. The absence of parity means write speeds are high. Along with excellent redundancy, this is probably the best option for speed and redundancy.
RAID 0+1	4	N/2
RAID 0 + 1 is not interchangeable with RAID 10. There is one huge difference and that is reliability. You can lose only one drive and have a functioning array. With the more drives in a single RAID 0 stripe the greater the chance you take. Speed characteristics are identical to RAID 10. I have never implemented RAID 0 + 1 when RAID 10 was available.
RAID 50	6	(N-1)*R
Since RAID 5 becomes more susceptible to failure with more drives in the array keeping the RAID 5 stripe small, usually under 8 drives and then striping them with RAID 0 increases the reliability while allowing you to expand capacity. You will lose a drive per RAID 5 stripe but that is a lot less than loosing half of them in a RAID 10. Before RAID 6, this was used to get higher reliability in very large arrays of disks.
RAID 60	8	(N-2)*R
RAID 60 is the exact same concept as RAID 50. Generally, a RAID 6 array is much less susceptible to an array failure during a rebuild of a failed drive due to the nature of the dual striping that it uses. It still is not bullet proof though the RAID 6 array sizes can be much larger before hitting the probability of a dual drive failure and then a failure during rebuild than RAID 5. I do not see many RAID 60 configurations outside of SAN internal striping schemes. You do lose twice as many drives worth of capacity as you do in a RAID 50 array.
RAID 100	8	N/2
RAID 100 is RAID 10 with and additional RAID 0 stripe. Bridging multiple drive enclosures is the most common use of RAID 10. It also reduces the number of logical drives you have to maintain at the OS level.

Speed, Fault Tolerance, or Capacity?

You can’t have your cake and eat it too. In the past, it was hard to justify the cost of RAID 10 unless you really needed speed and fault tolerance. RAID 5 was the default because in most situations it was good enough. Offering near raid 0 read speeds. If you had a heavy write workload, you took a penalty due to the parity stripe. RAID 6 suffers from this even more so with two parity stripes to deal with. Today, with the cost of drives coming down and the capacity going up RAID 10 should be the default configuration for everything.

Here is a breakdown of how each RAID level handles reads and writes in order of performance.

RAID Level	Write Operations	Notes	Read Operations	Notes
RAID 0	1 operation	High throughput, low CPU utilization. No data protection	1 operation	High throughput, low CPU utilization.
RAID 1	2 IOP’s	Only as fast as a single drive.	1 IOP	Two read schemes available. Read data from both drives, or data from the drive that returns it first. One is higher throughput the other is faster seek times.
RAID 5	4 IOP’s	Read-Modify-Write requires two reads and two writes per write request. Lower throughput higher CPU if the HBA doesn’t have a dedicated IO processor.	1 IOP	High throughput low CPU utilization normally, in a failed state performance falls dramatically due to parity calculation and any rebuild operations that are going on.
RAID 6	6 IOP’s	Read-Modify-Write requires three reads and three writes per write request. Do not use a software implementation if it is available.	1 IOP	High throughput low CPU utilization normally, in a failed state performance falls dramatically due to parity calculation and any rebuild operations that are going on.

Choosing your RAID level

This is not as easy as it should be. Between budgets, different storage types, and your requirements, any of the RAID levels could meet your needs. Let us work of off some base assumptions. Reliability is necessary, that rules out RAID 0 and probably RAID 0+1. Is the workload read or write intensive? A good rule of thumb is more than 10% reads go RAID 10. In addition, if write latency is a factor RAID 10 is the best choice. For read workloads, RAID 5 or RAID 6 will probably meet your needs just fine. One of the other things to take into consideration if you need lots of space RAID 5 or RAID 6 may meet your IO needs just through sheer number of disks. Take the number of disks divide by 4 for RAID 5 or 6 for RAID 6 then do your per disk IO calculations you may find that they do meet your IO requirements.

Separate IO types!

The type of IO, random or sequential, greatly affects your throughput. SQL Server has some fairly well documented IO information. One of the big ones folks overlook is keeping their log separate from their data files. I am not talking about all logs on one drive and all data on another, which buys you nothing. If you are going to do that you might as well put them all on one large volume and use every disk available. You are guaranteeing that all IO’s will be random. If you want to avoid this, you must separate your log files from data files AND each other! If the log file of a busy database is sharing with other log files, you reduce its IO throughput 3 fold and its data through put 10 to 20 fold.

RAID Reliability and Failures

Correlated Disk Failures

Disks from the same batch can suffer similar fate. Correlated disk failures can be due to a manufacturing defect that can affect a large number of drives. It can be very difficult to get a vendor to give you disks from different batches. Your best bet is to hedge against that and plan to structure your RAID arrays accordingly.

Error rates and Mean Time Between Failures

As hard disks get larger the chance for an uncorrectable and undetected read or write failure. On a desktop drive, that rate is 10^14 bits read there will be an unrecoverable error. A good example is an array with the latest two-terabyte SATA drives would hit this error on just one full pass of a 6 drive RAID 5 array. When this happens, it will trigger a rebuild event. The probability of hitting another failure during the rebuild is extremely high. Bianca Schroeder and Garth A. Gibson of Carnegie Mellon University have written an excellent paper on the subject. Read it, it will keep you up at night worrying about your current arrays. Enterprise class drives are supposed to protect against this. No study so far proves that out. That does not mean I am swapping out my SAS for SATA. Performance is still king. They do boast a much better error rate 10^16 or 100 times better. Is this number accurate or not is another question all together. Google also did a study on disk failure rates, Failure Trends in a Large Disk Drive Population. Google also found correlated disk failures among other things. This is necessary read as well. Eventually, RAID 5 just will not be an option, and RAID 6 will be where RAID 5 is today.

What RAID Does Not Do

RAID Doesn’t back your data up. You heard me. It is not a replacement for a real backup system. Write errors do occur.As database people we are aware of atomic operations, the concept of an all or nothing operation, and recovering from a failed transaction. People assume the file system and disk is also atomic, it isn’t. NTFS does have a transaction system now TxF I doubt SQL Server is using it. Disk drives limit data transfer guarantees to the sector size of the disk, 512 bytes. If you have the write cache enabled and suffer a power failure, it is possible to write part of the 8k block. If this happens, SQL Server will read new and old data from that page, which is now in an inconsistent state. This is not a disk failure. It wrote every 512-byte block it could successfully. When the disk drive comes back on line, the data on the disk is not corrupted at the sector level at all. If you have turned off torn page detection or page checksum because you believe it is a huge performance hit, turn it back on. Add more disks if you need the extra performance don’t put your data at risk.

Final Thoughts

Data files tend to be random reads and writes.
Log files have zero random reads and writes normally.
More than one active log on a drive equals random reads and writes.
Use Raid 1 for logs or RAID 10 if you need the space.
Use RAID 5 or RAID 6 for data files if capacity and read performance are more important than write speed.
The more disks you add to an array the greater chance you have for data loss.
Raid 5 offers very good reliability at small scale. Rule of thumb, more than 8 drives in a RAID 5 could be disastrous.
Raid 6 offers very good reliability at large scales. Rule of thumb, less than 9 drives you should consider RAID 5 instead.
Raid 10 offers excellent reliability at any scale but is susceptible to correlated disk failures.
The larger the disk drive capacity should adjust your number of disks down per array.
Turn on torn page for 2000 and checksum for 2005/08.
Restore Backups regularly,
RAID isn’t a backup solution.