Slashdot Mirror
Mirroring Controllers - What have been Your Experiences?
Today's installment is a lengthy (but hopefully informative) piece on mirroring controllers. Ever had weird problems with a FastTrack TX2000? Are you curious how well HighPoints RocketRAID boards are really supported? Ever wondered which controller gives you the best performance for every dollar spent? In true Slashdot tradition, we're taking the issue and throwing it out to you, the readers. Futurepower(R) is willing to start us off with a wealth of information on his experiences, and I'm hoping a few of you are willing to do the same.
Futurepower(R) asks: "What experiences, both good and bad, have you had with mirroring controllers? Are there manufacturers I haven't found?
For those who are interested but don't have experience, mirroring controllers (RAID 1) provide several advantages:
ECS (EliteGroup) has made 11 motherboards with on-board RAID mirroring controllers. One of them, the P4VXAD, has a Promise controller and costs about $52. What's that about? Employees at Fry's tell me that ECS has the poorest quality control of any of the motherboard manufacturers that they sell. To me, ECS motherboards seem to have surprisingly high quality. However, we have only tested three, and only one extensively.
Silicon Image makes, or made, their 0680 RAID chipset, and Koutech Systems sells the IOFLEX-Pir133 using that chipset. It costs $25 retail. I tried a Koutech card and had a lot of trouble with it, even after updating the BIOS. I talked to an application engineer at Silicon Image and a manager there. I was told that SI bought the technology from another company, and apparently it is no longer supported, doesn't work well, and the company does not intend to put more money into it. I found that the Koutech card corrupts files. The card gives a Stop 07B error when going into the Windows XP Recovery Console, unless the driver is loaded by hand, every time. I lost a lot of time with a product that apparently should not be sold.
My experience with mirroring controllers is mostly with those from Promise Technology. I've been using Promise mirroring controllers since they began making them.
The good about Promise:
The Acard controller supports SuSE, Red Hat, Caldera, and Turbo Linux, it says. But remember, Froogle found only one vendor.
HighPoint says they support Linux: "Linux Red Hat 7.3 & 8.0 (Software RAID Only)". This apparently means the card does not support Linux at all, since Linux has software RAID built in.
So, that's the extent of my knowledge and experience. Can you provide further insight?"
For those who are interested but don't have experience, mirroring controllers (RAID 1) provide several advantages:
- They prevent data loss when a hard drive fails. The other drive in the mirror takes over seamlessly.
- Reading of data is considerably faster since the controller reads the data from the drive that has a head closest to the data.
- You always have a full hard disk backup that you can pull from your system at any time, such as before installing new software.
- You can clone a Windows XP drive using the RAID card, and you will have a bootable copy. This is valuable, since the Windows XP file system cannot copy all of its own files. (Don't flame this; it has been verified many times by Microsoft employees, who often suggest using the third-party tools.)
- Promise Technology's FastTrak TX2000 is available for about $85 delivered.
- HighPoint's RocketRAID 133 costs about $80.
- ACARD Technology's AEC-6880 costs about about $85. Froogle found only one vendor. That's scary.
- Adaptec makes the ATA RAID 1200A, which is available for about $59 before delivery charges.
ECS (EliteGroup) has made 11 motherboards with on-board RAID mirroring controllers. One of them, the P4VXAD, has a Promise controller and costs about $52. What's that about? Employees at Fry's tell me that ECS has the poorest quality control of any of the motherboard manufacturers that they sell. To me, ECS motherboards seem to have surprisingly high quality. However, we have only tested three, and only one extensively.
Silicon Image makes, or made, their 0680 RAID chipset, and Koutech Systems sells the IOFLEX-Pir133 using that chipset. It costs $25 retail. I tried a Koutech card and had a lot of trouble with it, even after updating the BIOS. I talked to an application engineer at Silicon Image and a manager there. I was told that SI bought the technology from another company, and apparently it is no longer supported, doesn't work well, and the company does not intend to put more money into it. I found that the Koutech card corrupts files. The card gives a Stop 07B error when going into the Windows XP Recovery Console, unless the driver is loaded by hand, every time. I lost a lot of time with a product that apparently should not be sold.
My experience with mirroring controllers is mostly with those from Promise Technology. I've been using Promise mirroring controllers since they began making them.
The good about Promise:
- Promise has been selling RAID 1 controllers a long time. They work.
- Promise controllers can clone a hard drive quickly.
- Now some motherboards have Promise mirroring chipsets. You can get the entire motherboard with the Promise controller on the board, for maybe $50 more than the Promise controller card alone.
- Linux drivers are available. The web site says, "Windows XP/2000/NT4/Me/9x; Novell NetWare 4.1x/5.x; RedHat Linux 7.0/7.1/7.2; TurboLinux Server 6.5; TurboLinux Workstation 7; SuSE Linux 7.2; OpenLinux 3.1" Does that mean that later versions of Linux can't use this card, or did Promise forget to update the brochure PDF file?
- Promise controllers work fine under DOS, but there is no error reporting if for some reason the mirror breaks.
- Promise mirroring controllers have a software feature called "sychronization". I've asked many times over the years why it is necessary, since mirrored hard drives should be synchronized 100% of the time. I've never gotten an answer. Recently I've been told by Promise technical support people not to use synchronization, since it has caused problems. It sounds like some technical problem is being hidden.
- Promise does not support their oldest mirroring controllers under Windows XP. This is a problem since there are many business computers that are used for data entry. A Pentium II is as fast as is necessary. Windows 98 is stable with only one program running. Now those computers need to be converted to Windows XP, since Microsoft has declared that its operating systems have a curious quality: They die. (According to Microsoft, it doesn't matter that at least 100,000,000 people are using Windows 98 worldwide, it came to the end of its life on "30-Jun-2003".)
So, it is necessary to buy another controller for old data entry systems. Notice that Microsoft and Promise could decide to play this game again, and I would like to avoid the second round of buying and installing even another controller. I'd like to find a company that continues to support its products.
The speed of computers used for data entry does not matter, but the security of the data does. Hard drive failures are becoming rare, but a hard drive failure can cause a lot of problems on a data entry computer, so mirroring is required. - Some Promise controllers, especially those on motherboards, take a long time to boot. Dots crawl across the screen even if no drives are connected to the controller. Is keeping the Promise name on the screen a time-wasting sales message from Promise? Recently Promise released a BIOS upgrade for some of its cards that reduced the dot-crawling time. However, there is apparently no upgrade for Promise controllers on motherboards.
- About 2 months after I reported problems, Executive Software said they found a bug in their Diskeeper defragmentation software that might cause data corruption when used with Promise controllers. They said everyone using Diskeeper should upgrade to the new, free, minor version. I've seen no problems since then.
- Promise Technology's sales literature can be disgusting. This is the second sentence in Promise's description of the FastTrak TX2000: "The FastTrak TX2000 ATA RAID card supports Ultra ATA/133 drives to rock workstations and boost small (or large) office servers like never before." To me, this is obviously written by someone who knows nothing about the product and doesn't care.
- I find the abundant use of PDF files and unnecessary JavaScript on Promise Technology's web site annoying.
The Acard controller supports SuSE, Red Hat, Caldera, and Turbo Linux, it says. But remember, Froogle found only one vendor.
HighPoint says they support Linux: "Linux Red Hat 7.3 & 8.0 (Software RAID Only)". This apparently means the card does not support Linux at all, since Linux has software RAID built in.
So, that's the extent of my knowledge and experience. Can you provide further insight?"
I've used the Adaptec card before, and I learned the hard way that you need to be aware of heat. Two modern high-speed IDE drives generate a large amount of heat, and if you mount the two directly next to one another in the case you may well burn one out, as I did.
The dead drive may have had some other defect, but I doubt it as the thermal alarm on the MB would sound after a few hours of usage. Even after replacing the drive (and being sure the two drives were mounted far enough apart to allow airflow), the machine still occasionally comes up with blank disks, requiring an OS reinstall.
Just something to keep in mind when you're building your system--YMMV.
usually (afaik/iirc/if_i_got_a_clue) the cards that are cheap use the cpu on motherboard for the raid functionality, and as far as i'm concerned i find their 'only' use to be cheap(ish) way to add more ata drives to a system.
get a real card or you might just as well use software raid, the good models don't come with 64mb of mem just for kicks either, and the good models have linux support too (adaptec 2400a, if that was the model i installed once, i think it is, even came with managing software that was a bootable linux cd). sure it's expensive but it's REAL, not just an ata interface and programs.
world was created 5 seconds before this post as it is.
Added Information:
I wrote the Slashdot story. I've gotten additional information since then:
First, I don't see any evidence of low quality in ECS motherboards. Both ECS and Fry's employees say that the high rate of returns is due to the fact that the low cost attracts people who are very inexperienced at building computers. That seems right to me. An ECS sales representative mentions that Fry's uses ECS motherboards in the computers it builds. The ECS motherboards are suitable for business use. They may not be the best for overclockers or gamers.
I wrote a long letter to Promise Technology Technical Support about the fact that Promise is allowing ECS motherboards with Promise RAID mirroring controlers to be sold without software that is necessary to monitor the health of the mirror. The letter is below, and represents my opinions at the time. Even though the national sales manager of Promise suggested I send the letter, there was no answer.
Below the letter is all information and opinions that I had available at the time.
___________________
TO:
Kevin Hong
Promise Technology Technical Support Engineer
Kevin,
Jason Turk told me to write to you about a problem with Elitegroup's L7VTA Rev. 1.0 motherboard, which includes a Promise RAID controller chip.
Elitegroup says they have never been provided the FastCheck or PAM monitoring software for this motherboard. Can you supply monitoring software that is certified by Promise to work? The software is not on the CD supplied with the motherboard.
We need not just software that appears to work, we need software that is certified by Promise Technology to work with this motherboard. Several people at Promise have told me that the software provided on the Promise web site for the Promise RAID controller adapter cards may not work with the OEM controller chips installed on motherboards. I suspect this is not true, but merely a Promise Technology marketing scheme. However, we can't risk selling these to our customers and finding later that there is some hidden serious defect.
Thanks,
Michael Jennings
Futurepower Computer Systems
_______________________________
Complete Information:
So that all the information to resolve this issue is supplied in one place, I have provided everything that I know and think below:
Some versions of the Elitegroup L7VTA Rev. 1.0 motherboards are being supplied with Promise Technology's PDC20265R IDE RAID controller. Elitegroup supplies three of the four elements needed for a RAID controller. The motherboard and included CD have 1) Promise Technology's PDC20265R ASIC chip, 2) the latest version of the BIOS, 3) and the driver software.
However, item 4 is missing. Elitegroup says Promise Technology never supplied RAID array monitoring software for this chip. Promise Technology calls this software the "FastCheck" or the "PAM, Promise Array Management" utility. Without this software, the user cannot know if a hard drive has failed in the RAID 0 or RAID 1 or RAID 0,1 array.
This issue is Elitegroup's Case number RAE54616. (Note that Elitegroup also calls themselves "ECS".)
Elitegroup's web page for the L7VTA product is here:
http://www.ecs.com.tw/products/pd_spec.asp?product _id=327
The L7VTA V1.0 driver page is here, showing the latest Promise drivers:
http://www.ecs.com.tw/download/dw_spec.asp?product _id=63
Promise Technology's web page for on-motherboard RAID controllers is here:
http://www.promise.com/product/oem_ataraid_pdc2026 5r_eng.htm
Note that this Promise web page mentions all four elements of a
On the other end of the spectrum, however, is OS-level mirroring. I have only experience with Linux RAID-1, both that has never failed me either. I have run software RAID-1 for many years and have only (long ago) had problems where I had to "raidhotadd" a drive from a mirror set that was no longer recognized.
My recommendation, based on both performance and ease of use, is to use a 3ware controller. That coupled with Linux LVM makes disk management and resiliency a breeze.
My only disappointment with the 3ware controller is that one cannot make take a single disk with data on it, add a second disk and tell the controller to mirror the first disk to the second. With the 3ware, you will lose the data on both disks when making a mirror. (So says 3ware tech support.)
the growth in cynicism and rebellion has not been without cause
Raid 1 basically issues a read request to both drives and uses the data returned from whichever drive responds first.
Does RAID 1 specifically require that, or would that count as an implementation-specific detail?
Because, if implementation-specific, I see no reason why RAID 1 couldn't use the same striped request pattern, since the desired data will exist on both drives anyway (ie, a proper superset of the data available to RAID 0). That wouldn't speed up writing, but for reading, you could get the same benefit as RAID 0.
They may be great products, but they are a bit expensive for home use.
The prices can be a bit difficult to find using the site, but all serial ATA raid controllers were above 400 US dollars.
I've heard of these controllers before, and people seem to like them very much. But at this price-point (and maybe availability too) they seem to be aimed more at the professional market.
Which is fine if you are a professional user, obviously.
Warper
Software RAID (under Linux) is a better deal than spending $60 on a RAID card. The RAID drivers for these cards basically just implement software RAID anyway, so what's the point of having the hardware? The card is just one more thing to go wrong. Use the IDE controller that comes with your mobo.
Software RAID has many advantages. You can monitor everything about the RAID array from within Linux. Drivers are no problem. It is well-tested and stable.
Also, you can get better performance. I would rather spend an extra $200 for a dual-CPU mobo, get a second CPU, and run software RAID, than spend $200 to get a dedicated RAID card.
Unless I'm mistaken these cards don't do RAID anything. Indeed, they are ordinary controllers that have a few extra functions, all of which are basically hardware acceleration for RAID operations. The critical bits of the RAID are still done in software.
With that said, if you're going to do software RAID, you might as well just do it. I've been using VINUM on FreeBSD for a long time, there are other options for FreeBSD and for Linux. The better ones take full advantage of the "dual read" property of mirrored drives at the kernel level. Best part, no special hardware required. If something breaks buy any old ATA card and you're back in business since it's all in software.
If you need RAID-4 or RAID-5, a dedicated card might be a good idea. For mirroring I think software is superior.
...I've been running it for 6 months under Mandrake 9.1. Smooth sailing...I've had RAID 0 and 1 arrays on it without any problems. Also supports 1+0 and JBOD. The RR454 is out now, too, which supports RAID 5 (ooh, if only I'd waited...) Great value, though--can run 8 drives for ~90 bucks. Hit up pricewatch or something, they're easy to find.
I think this is due to the way 3ware formats the disks for use in an array. You can't just pull a disk out of an array (even if it was in JBOD or RAID1 mode0), slap it on any old spare drive controller, and read the data - the format is specific to a 3ware controller.
That said, any 3ware controller can read a disk that was written to by another controller (the format hasn't changed), provided the physical interface is the same. I've used drives from a 6400 on a 7500 without any trouble.
The benefits of the special format basically amount to being able to identify what part of the array a drive is from, and it's function, while swapping them around. You can plug disk 3 into disk 2's connector, and the card will identify and deal with the situation appropriately. Pretty useful in a larger array, or if you use the same hotswap cartridges on multiple systems, and get them mixed up.
I bought a 3Ware 6400 back when they were new. Before the 6410 came out, and you had to get a full length PCI card for 4 ports. I got it through hypermicro.com, who I knew and trusted for all my storage purposes.
Until the 3ware, I had been "The SCSI guy" in most circles. I was running all scsi systems starting in 1992 and woe betide he who suggest I add an inferior IDE device to my system. I had first spotted the 3Ware at PC Expo back when they were prototyping the 6000 series, and immediately scoffed at it as "just another IDE RAID wannabe." The 3Ware engineers overheard this and invited me to come over and play with their system and tune the benchmarks on it (standard ones like iometer, not custom in house things you can fake easy) to prove they were serious. I left their booth saying "I want one!" and repeating it to my friends for the rest of the show.
But I digress. I had purchased a 3ware, to go in my new Athlon system. This was back in the old Slot A athlon days with the bad AMD750 chipsets. I had an Asus K7M, and the 3Ware was not cooperating. After a few hours of fighting with it, I put it in my trusty celeron system with the BX chipset motherboard. It still didn't work, so then I knew I had a problem. This was unfortunate considering I was leaving for college on saturday, and this was thursday afternoon.
So I called 3Ware. I described the problem and had very little trouble (for once) convincing the tech support guy I wasn't some kind of moron. They believed me when I said I knew what I was doing when I put a system together, unlike most other companies. After an attempt to flash the card's BIOS to a good image, they said that the card most likely had a defective chip and would need replacement. They said they would gladly do it but their RMA system was down due to a server problem, and that calling hypermicro back would be the best option.
So I called Hypermicro, and they say they'll be happy to get me a new card, but there's no way they could get it to me overnight (as it's now late afternoon). I explain the problem of leaving for college and they say they'll be happy to ship it to me when I get there. I wasn't happy about that.
So I called 3Ware back, and explained what Hypermicro had said about being unable to replace the card before I leave. It's now about 8:30pm NY time, Hypermicro I believe is in central time, and 3Ware is on pacific time. The 3Ware support guy brings his manager on the phone (I didn't ask), and the manager says "Give me your information, and I'll have a card there for you. Send us your defective one back when you get a chance." So, 3Ware sends me a retail box card with the cables and all as a replacement for my OEM bare card, cross shipped, without an RMA number, and has someone take a car and speed to the FedEx hub in person to get it shipped.
10AM NY time the next morning the card was there.
The new card worked perfectly, and I was very happy with it. About 2 months after I bought it, 3Ware sends me an email. It basicly says the following: "We have implemented RAID5 support for your card in our new driver and firmware. You may download it as a free upgrade even though we had only planned to support RAID0, 1, and 10 on your product." How's that for support and driver updates?
Also, after the 6000 series was retired in favor of the 7000 and 7500s, 3Ware sends me another email that basicly says "We know your cards are considered obsolete and no longer made, but here's a free driver that adds > 137GB disk support to your ata-66 level controller that shouldn't officially support that standard." And at that point they'd basicly retired the products, yet they were still developing new drivers and firmware for their obsolete parts!
I now own a 7500-4 as my main controller, and I will likely be buying one of the 9000 series SATA-II controllers when they come out as an upgrade.
Oh and I've never missed my DPT (unfortunately since bought and ruined by Acraptec) with its hardware RAID5 and cache or my seagate cheetahs or any of my other SCSI hardware ever since.
Introducing the new Occam Fusion! Now with sqrt(-1) fewer blades!
The array switches to Disabled state at random, with no apparent cause. This happened twice yesterday, but not at all for the previous 8 days. I contacted Highpoint support and their advice was (1) make sure the BIOS version matches, and (2) check cables and connectors.
I'm using a beta version of their opensource driver which is not linked from their webpage. I put it online at hpt374-opensource-v2.01-0718.tgz for any others who need v2 driver source rather than v1.
I'm running 2.4.21 with LVM. I did have the root filesystem on LVM under the Highpoint controller but had problems with extending it. So it is now under the inbuilt IDE controller, using software RAID and no LVM. I find my box is less fiddly to configure if the root filesystem is not LVM.
When the drivers set the array state to "Disabled" no more disk I/O is possible to those devices. A reboot is required. Strangely, the devices have never required resyncing after this problem.
At the subsequent reboot linux usually spends a few seconds replaying the filesystem journal (they're all EXT3 filesystems).
I have a theory that the Highpoint card (or driver) does not like sharing interrupts. Since the problems yesterday I have fiddled with the IRQ assignment and one problem that I had noticed (occasional ghost streaks on the CRT display, signifying contention on the video memory during a raster trace interval) has apparently disappeared. It may be that the Highpoint controller was sharing an interrupt with the display controller, making one of them not very happy. Thus, I'm cautiously optimistic that my problems may be cured.
In summary I can't really recommend the highpoint card for stable use. It's great to get a lot of IDE channels (e.g. using the quad-channel ATA card, and putting 2 devices per channel) and SATA is great for cabling drives, but there are obviously some driver issues which need to be worked out first.