Comparison of Nine SATA RAID 5 Adapters

Robbedoeske writes "Tweakers.net has put online a comparison of nine Serial ATA RAID 5 adapters. Can the establishment counter the attack of the newcomers? Which of the contestants delivers the best performance, offers the best value for money and has the best featureset?"

Comments

Eight or Nine?

[SirTwitchALot]

TFA says nine adapters, but the graphic says eight, whoops.

Re:Eight or Nine?

[pbranes]

They must have started counting at 0. Stupid off-by-1 errors. ;-)

Seriously, though, I have been seeing many servers start to come in with SATA drives. Right now it is low end and off-brand servers. Dell even ships SATA drives in their cheapest server line. Sure SCSI has high spin rates & throughput, but they are freakin expensive. A good SCSI raid controller costs close to $1000 and a good SCSI hard drive can cost $400. It is so expensive, that it is reallly worth it sometimes to get the SATA drives in servers. I haven't seen that reliability of SATA over SCSI is a problem. I'm truly hoping that SCSI goes the way of the dodo. Its a pain to use. Who know what kind of cable you're supposed to use with that external SCSI device. SCSI, in its current form, is just opening itself up to becoming antiquated.

Re:Eight or Nine?

[ackthpt]

TFA says nine adapters, but the graphic says eight, whoops.

It was a parity bit, ignore it.

32 pages? No thanks.

[Mr. Sketch]

After 32 pages, it's probably just best to skip to the conclusion:
http://www.tweakers.net/reviews/557/32

Where it has the executive summary:

Areca ARC-1120: highly recommended
RAIDCore BC4852: recommended
HighPoint RocketRAID 1820A: recommended

For several reasons, we will refuse recommendations on the remaing adapters in this comparison

I think that pretty much covers the jist of the article.

Interesting that the 3ware offerings performed...

[tabkey12]

quite badly! They've been synonymous with quality in the RAID industry for many years. Look at this:

3ware Escalade 8506-8 is lagging far behind the competition. Moreover, it misses important features such as online capacity expansion, online RAID level migration and RAID 50 support.

http://www.tweakers.net/reviews/557/6

What they say in the article is almost damning really...

Re:32 pages? No thanks.

[tabkey12]

But RAID is nota one size fits all game - the detail of the article is extremely useful for people who will be tailoring their RAID to a specific application. Yes, this article is specialised, but I hardly see how reducing it to a list of three, relatively meaningless names is helping.

Re:3ware me

[B3ryllium]

The previous poster had this to say:

3ware Escalade 8506-8 is lagging far behind the competition. Moreover, it misses important features such as online capacity expansion, online RAID level migration and RAID 50 support.

Don't plan on mixing Highpoint cards

[GoodNicsTken]

I had a Rocket Raid 100 (IDE 4 drive RAID1/0) and a RocketRaid 1640 (4 Channel SATA RAID 0,1,5) card. With nothing connected to the 1640 and 2 mirrored drives on the RR 100 the disks attached to the RR100 in bios show up on the 1640, and when windows gets to the boot screen it locks up.

When I removed the drives in windows, it booted up without problems. Highpoint has sent me diag tools to run rather than building this in their lab!

I'm not too impressed with them so far.

You know the cheap-reliable-fast triangle.

[lanc]

Well, cheap+reliable == linux + softraid + Enhanced Network Block Device + Enterprise Volume Management System (or LVM2). It is often faster than non-hw-raid (fake-hw controllers.

Drivers?

[sjbe]

While I've admittedly not read the entire article (it's really long) I couldn't find much info about drivers. It seems the author basically assumed one would be running windows, which for servers (the most likely place for a RAID array) is a pretty poor assumption. I've tried a number of SATA RAID cards on my linux server (SuSE 9.1) and keep getting driven back to SCSI due to crappy/non-existant driver issues. Thank god for Addonics SATA-SCSI adaptors which work great and have saved me a bunch of money.

It's a nice article comparing performance but without a serious analysis of drivers along with it for Windows AND linux (and Mac if applicable) the article isn't complete. I don't really care which one is fastest if I can't run it on my system.

Re:Drivers?

[LWATCDR]

3-ware has very good support for linux

Re:Interesting that the 3ware offerings performed.

[arivanov]

Well... as someone who has both of reviewed 3ware adapters in production I am not amazed. They are nice, but nothing to shout about. They also have LOADS of PROBLEMS not mentioned in the article.

• 8506 SATA series prior to a certain board revision are extremely susceptible to bus noise. As a result you have to find a way to bastardize the PCI bus down to 33MHz and provide additional grounding. Even so, they are likely to cause random system deaths and serious memory corruption in most Opteron MSI and Assus motherboards as well as some other designs. Using in 1U and 2U chassis with riser cards is a no-no for the same reason (exemption for some buffered risers). As a side note, most resellers will try to stuff you with an old board despite the fact that they know about this problem.
• 9506 board and linux driver at least as of 2.6.9 defaults to no read cache, only write cache which is outright daft. It is also the major reason for low performance at least under Linux.

Both are nice cards, but I would not recommend them to anyone who does not have extensive PC hardware knowledge. They are fussy, carpicious and very hard to troubleshoot when they go wrong.

If it doesn't say

[cmefford]

3ware, It's a waste of /my/ time. So I really don't care what the tweakers think. The 3ware cards are reliable, easy to deal with, have brilliant drivers, good software, and they WORK! Always! I have a 5 of them. I have a friend who has 40, he agrees. I use a 2 channel as a backup-to-sata drive, (cheaper than tape), another 2 channel in a IIS server for payroll stuff, 1 4 channel for a mail server and 2 8 channels for file/web. I love'em. Nuff said.

My thoughts

[tonsofpcs]

Areca ARC-1120 looks better on each and every page except for the sequential read/write tests where it tends to come in third [I'm just reading off the graphs].
The RAIDCore BC4852 seems fastest for sequential reads/writes.

BOTH of these have linux support. The Areca supports: Mandrake (9.0),Red Hat (7.3, 8.0, 9.0, AS 3.0), Fedora Core (2, 2 AMD64), SuSE (7.3, 9.1 Pro, 9.0 SLES, 9.0 SLES AMD64)
The RAIDCore: Red Hat (9.0, AS 3.0), Fedora Core (1)
The Areca also supports Windows XP and Server 2003 64-bit versions and BSDs: 4.2R, 4.4R, 5.2.1 (incl. source).

Also, the Areca ARC-1160 (they finished testing after the original article was written, so it didn't make it into most of the text) appears at the top of all of the Index/performance tests, except for "Fileserver - Large Filesize - RAID 1/10" and "My SQL - Data Drive - RAID 1/10".

SCSI vs SATA

[sjbe]

SCSI, in its current form, is just opening itself up to becoming antiquated.

Perhaps, though personally I've had far more trouble getting SATA (and IDE) drives to work than SCSI drives and I've used both extensively. Driver issues mostly. SCSI's performance is better in multi-user systems, it's easy to set up, drivers tend to be less problematic especially on systems other than Windows, and it can have more devices attached. People claim it's more reliable though I have no evidence of this, and frankly am a bit dubious of the claim. SATA is also easy to set up and is a lot cheaper, though the drivers are still less ubiquitous than with SCSI and performance doesn't match SCSI yet for multi-user systems. (on a single user system it doesn't matter much)

That said, the next generation of SCSI is Serial Attached SCSI which is compatible with SATA. A SAS controller will be able to use SATA drives if you don't need the extra features of SAS. SCSI isn't going away, it's just adapting.

Re:SCSI vs SATA

[gmezero]

Here's how I seperate SCSI vs SATA. I use SATA RAID setups for video workstations that need large drive space and cheap drives... and I don't care if the drive pops after a year of abuse.

I put SCSI in my servers (RAID or otherwise) when I want the box to run for years and years under heavy load and not have to worry about replacing drives regularly.

With SCSI, your paying for the quality control/quality assurance more than anything else.

From what I understand a good SATA drive has the same TTL quality as a good IDE drive, just faster performance.

My experience with 3ware

[dfn5]

I can say from my own experience with 3ware is that it sucks. We decided that we wanted to use S-ATA because we could get a lot of disk cheap. The problem was that these escalade cards didn't do parallel IO very well and by that I mean if one user is doing a long write operation the entire RAID array would go unresponsive to other users. For example if I created a large 20G oracle datafile the entire system would seem unresponsive until the operation completed. I wouldn't even be able to ssh into the server. And this was RedHat AS in case anyone wanted to know.

Moral of this story? You get what you pay for. SCSI should be used for servers.

To be fair, however, I was never able to determine if it was a result of using S-ATA, 3Ware or the linux device driver.

waste of time and money.

[Anonymous Coward]

I am running RAID 5 in my computer right now.

Linux software RAID. Makes all this crap obsolete except for some specific cases.

I can have as many drives as I want, I can have hot swapability, I can have hot spares and all sorts of fun stuff.

Add LVM on top of that and you have a solution that is much superior then going out and buying any raid controller, except for the most fastest.

Linux software raid is actually VERY nice, I don't know of any OS that has better setup.

Re:waste of time and money.

[Phil Wherry]

So, you're saying that somehow your software RAID is calculating XOR bits and such without putting a serious hurt on your CPU and memory? Interesting.

I'm not the original poster to whom you're responding, but the answer to your question is actually kind of surprising (it was to me, in any case).

At any given point in time, your system is in one of three states:

• partially idle (there's unused CPU and disk I/O capacity),
• CPU-bound (the CPU is fully utilized but there's disk I/O bandwidth available), or
• I/O bound (the CPU has spare cycles, but the disk can't provide data fast enough to put them to use).

I suppose that a heavily overloaded system could be both CPU and I/O bound, but it would require a mix of CPU- and disk-intensive processes that isn't usually seen in practice.

Let's ignore the partially idle case, in which there's ample disk and CPU to go around, as it doesn't really matter in this scenario whether the CPU or disk controller perform the XOR operations.

In the case of a CPU-bound process, you're going to incur the additional CPU overhead of the XOR operation. XOR is almost absurdly fast, particularly if the data is in the CPU's cache. I'm pretty sure that modern CPUs execute XOR on at least one byte per clock cycle. But let's say, for the sake of argument, that it takes three cycles per byte. On a CPU clocked at 3 GHz, you'd be able to perform XORs on one gigabyte of data per second if you ignore memory and cache issues. Given moderate memory bandwidth, you're also able to transfer over a gigabyte of data to or from the CPU per second. Given a more reasonable amount of data (say, one megabyte, to transfer), you'd be looking at a CPU impact of around one millisecond to perform the XOR. That's a 0.1% impact at most in a CPU-bound environment, and that's presuming you're doing a megabyte of disk I/O per second.

Now let's look at the I/O-bound case. Here, the CPU is sitting around waiting for the disk I/O to finish up. In this case, it clearly doesn't matter who's doing the XOR operations, since the CPU isn't fully utilized. PCI bus utilization is going to be increased by up to 100% (in the worst-case scenario involving drive mirroring; the worst-case RAID5 scenario is a 50% increase). A typical server's 66 MHz 64-bit PCI bus has a capacity of around 533 megabytes per second (PCI Express increases this dramatically, but let's stick with pessimistic examples for now). At the moment, a SCSI bus tops out at 320 megabytes per second, and those transfer rates are only achievable with at least four drives on the channel and an almost exclusively sequential I/O mix (the best-case numbers for a 15,000-RPM drive are about 100 megabytes/second). So there's generally bus bandwidth to spare.

You raise a number of other points in your note that are potentially issues (hot swappability, for example). But I've become convinced that the CPU/machine performance argument against software RAID really only made sense when CPUs/memory/bus bandwidth were much more constrained.

Beware hardware RAID

[puke76]

Sure everyone buys a few spare drives.. but make sure you buy more than one RAID card. If the RAID card goes, unless you replace it with an identical make and model, you can kiss your data goodbye.

That's what I like about software RAID on Linux - you can mount the array on another linux box if you need to.

Have yet to see a good comparison between low-end hardware RAID and Linux software RAID..

Re:Beware hardware RAID

[justins]

If the RAID card goes, unless you replace it with an identical make and model, you can kiss your data goodbye.

If you are dumb enough to use RAID as a substitute for backing up, that is.

The Article Never Explains What RAID 5 Is

[windowpain]

Now you could argue that a car review in Car and Driver doesn't bother explaining what a transmission does but RAID is several orders for magnitude more complex and esoteric.

There are so many different flavors of RAID it can be hard to keep them straight if you're not working with them every day.

Anyway there are good explanations of RAID here and here.

Re:The Article Never Explains What RAID 5 Is

[justins]

Now you could argue that a car review in Car and Driver doesn't bother explaining what a transmission does but RAID is several orders for magnitude more complex and esoteric.

Are you kidding?

RAID 5 can be explained in a few pages - the math, the implementation, the whole bit. Have you ever seen a technical drawing of a transmission? Modern slushboxes are about the most advanced mechanical engineering application that the average person ever comes in contact with (when they aren't at the airport).

You won't find an article that does most of the issues involved in designing and implementing a transmission justice. I know you just meant it as an example, but still. :)

Re:Comparison of Nine SATA RAID 5 Adapters

[killmenow]

It means that LINUX does NOT support them.

How you got modded Insightful is beyond me. Are you a troll? Did you RTFA? I didn't read it all either. But I read enough to catch this much: LINUX does support ALL of the cards in this review. Two of them (the Promise and RaidCore cards) come only in binary packages but the rest have Linux source available.

I, personally, would completely avoid any card manufactured by Promise or Highpoint as I've had crap luck with them in the past. They're just not very good cards, imho. And I'm not talking about their performance in Linux. I'm talking their performance in general. They're crap by my estimation regardless of platform. After losing data on my Windows 2000 box becuase of a crappy Highpoint card, I'll never buy another.

Anyway, your assertions are not even germaine. You point to the problem with "trick-BIOS" software RAID cards, which have been around for years and are not exclusive to SATA-RAID. They are shit cards, period...have been from the day they were made. Most of the cards in this review, however, are true hardware based SATA-RAID cards.

And, again, they all are supported on Linux. 3Ware, for example, has been a bastion of Linux support for ages.

As for the whole winmodem issue, who cares? What has it to do with a freaking troll blathering incorrectly about Linux not supporting SATA-RAID cards? Besides, the fact is, winmodems are NOT real modems. They're telecom interfaces, but not modems. You need software to make them modems. And I'm not talking about driver software to give access to the cards' functions. I'm talking software that has to implement the modem functionality itself...because the modem functionality doesn't exist on the "winmodem"...because it's not really a modem. Just because we now have linmodems.org and such to provide that software, it doesn't automagically make them "real" modems.