View previous topic :: View next topic |
Which SATA controller should I purchase |
AVLAB ( SiI 3112A ) |
|
20% |
[ 1 ] |
Highpoint RAID 1520 |
|
0% |
[ 0 ] |
Promise FastTrak TX2300 |
|
20% |
[ 1 ] |
ST Lab A-223 ( SiI 3114 ) |
|
60% |
[ 3 ] |
|
Total Votes : 5 |
|
Author |
Message |
theJackalnz n00b
Joined: 05 Oct 2005 Posts: 32
|
Posted: Thu Jun 29, 2006 6:45 am Post subject: Which SATA controller should I purchace |
|
|
PS: Thanks for those who voted in my poll, but please could I have comments to help validate your choices?
I have tried researching this myself, but I can't seem to find any solid usefull information.
I have four options.
- AVLab SATA Raid Controller, 2 Channel, PCI ( Silicon Image 3112A )[link] @ $61.70NZ
- Highpoint Rocket RAID1520, 2 Channel, SATA-150 RAID Controller, PCI[link] @ $166.28NZ
- Promise FastTrak TX2300, SATA2-300 RAID Controller, 2 Port, PCI[link] @ $175.62NZ
- ST Lab A-223, 4 Channel, 4x Internal, 2x External, SATA-150 RAID Controller, PCI ( Silicon Image 3114) [link] @ $68.68NZ
My general concerns are
Do they function well in linux ( ie: no bugs )
Do they have blatent hardware issues.
I dont want a crap drive, but I wan't to go as cheap as is "functionally affordable" ie: a wheels a wheel, nothing to be gained by having gold foil on the rims.
All the information I can find on the silicon image chips are bad, but also grossly out of date. Haven't seen anything newer than 2003 + linux 2.4 kernels.
And my mate reccons highpoint are a no-go with linux.
Recently, one of my SATA ( a seagate ) drives on my system has been giving me a lot of flack, to the extent where ive had to unplug it ( and all my home dirs , .. and portage ) just to make the machine operable
You could be working along and suddenly get :
Code: |
Jun 29 17:29:05 isengard ata2: translated ATA stat/err 0xd0/00 to SCSI SK/ASC/ASCQ 0xb/47/00
Jun 29 17:29:05 isengard ata2: status=0xd0 { Busy }
Jun 29 17:29:05 isengard sd 1:0:0:0: SCSI error: return code = 0x8000002
Jun 29 17:29:05 isengard sdb: Current: sense key=0xb
Jun 29 17:29:05 isengard ASC=0x47 ASCQ=0x0
Jun 29 17:29:05 isengard Info fld=0xa934392
Jun 29 17:29:05 isengard end_request: I/O error, dev sdb, sector 40004346
Jun 29 17:29:05 isengard ReiserFS: sdb5: warning: zam-7001: io error in reiserfs_find_entry
|
Which basically makes the entire drive inoprable, and all software that even accesses a SINGLE file on that drive ( all of them, my home dirs are on that drive ) Lock solid. Like a rock. the only solution that works, is a MAGIC-SYSRQ Sync , MAGIC-SYSRQ reBoot ( I dont think the Sync is doing me any good, but I do It any way just in case )
This is TORTURE, sometimes you can't even BOOT the machine without it happening, other time it can go for WEEKS without happening.
After dumping my messages files and cleaning and sorting the sector numbers, they seem to occur all in the same place, despite mutiple boots, So will be attempting to get the drive under waranty replacement.
BUT i have to stick another hard drive in so I can get all my data off, and Do do that, i need a controller.
So please, save me from my agony, and tell me whats crap |
|
Back to top |
|
|
NeddySeagoon Administrator
Joined: 05 Jul 2003 Posts: 54578 Location: 56N 3W
|
Posted: Thu Jun 29, 2006 11:46 am Post subject: |
|
|
theJackalnz,
Both the Silicon Image chipsets work with linux.
The 3112 has issues with some drives. The kernel maintains a 'blacklist'. Blacklisted drives will not operate at full speed with the 3112.
Its done to avoid data corruption.
The highpoint and Promise both have issues.
Check your old drive on the new controller before you return it. Maybe your onboard SATA is dieing? _________________ Regards,
NeddySeagoon
Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail. |
|
Back to top |
|
|
theJackalnz n00b
Joined: 05 Oct 2005 Posts: 32
|
Posted: Thu Jun 29, 2006 3:31 pm Post subject: |
|
|
Thanks for the reply, neddy.
I have 2 SATA drives on the current controller, and only one is seeming to have the problem.
It appears only sportadically, but causes havoc when it does.
Last time this blighter blew up on me, i fried my reiserfs partition quitebadly, and it got even more toasted when I tried to do the full-tree recovery on it.
( word of warning from somebody whos been there: dont store reiserfs formatted disk images/dumps on a reiser fs partition and perform a full tree rebuild. Reiserfs is too smart and will unpack those images into your FS for you )
Been doing a bit of SMART analysis on the drive,
Seagate - Purchased April 2005 ( warrany expires april 2010 luckily enough )
Code: |
=== START OF READ SMART DATA SECTION === ( SEAGATE drive. Current age: 1 year )
SMART Attributes Data Structure revision number: 10
Vendor Specific SMART Attributes with Thresholds:
ID# ATTRIBUTE_NAME FLAG VALUE WORST THRESH TYPE UPDATED WHEN_FAILED RAW_VALUE
1 Raw_Read_Error_Rate 0x000f 053 047 006 Pre-fail Always - 55863328
3 Spin_Up_Time 0x0003 098 096 000 Pre-fail Always - 0
4 Start_Stop_Count 0x0032 100 100 020 Old_age Always - 64
5 Reallocated_Sector_Ct 0x0033 098 098 036 Pre-fail Always - 87
7 Seek_Error_Rate 0x000f 085 060 030 Pre-fail Always - 357431310
9 Power_On_Hours 0x0032 091 091 000 Old_age Always - 8416
10 Spin_Retry_Count 0x0013 100 100 097 Pre-fail Always - 0
12 Power_Cycle_Count 0x0032 100 100 020 Old_age Always - 72
194 Temperature_Celsius 0x0022 029 053 000 Old_age Always - 29
195 Hardware_ECC_Recovered 0x001a 053 047 000 Old_age Always - 55863328
197 Current_Pending_Sector 0x0012 100 100 000 Old_age Always - 0
198 Offline_Uncorrectable 0x0010 100 100 000 Old_age Offline - 0
199 UDMA_CRC_Error_Count 0x003e 200 110 000 Old_age Always - 693
200 Multi_Zone_Error_Rate 0x0000 100 253 000 Old_age Offline - 0
202 TA_Increase_Count 0x0032 099 252 000 Old_age Always - 1
|
and my Hitachi Deskstar ( on the same controller ) I brought in September 2004
Code: |
ID# ATTRIBUTE_NAME FLAG VALUE WORST THRESH TYPE UPDATED WHEN_FAILED RAW_VALUE
1 Raw_Read_Error_Rate 0x000b 095 095 060 Pre-fail Always - 393221
2 Throughput_Performance 0x0005 100 100 050 Pre-fail Offline - 0
3 Spin_Up_Time 0x0007 182 182 024 Pre-fail Always - 177 (Average 152)
4 Start_Stop_Count 0x0012 100 100 000 Old_age Always - 249
5 Reallocated_Sector_Ct 0x0033 100 100 005 Pre-fail Always - 5
7 Seek_Error_Rate 0x000b 100 100 067 Pre-fail Always - 0
8 Seek_Time_Performance 0x0005 100 100 020 Pre-fail Offline - 0
9 Power_On_Hours 0x0012 098 098 000 Old_age Always - 14470
10 Spin_Retry_Count 0x0013 100 100 060 Pre-fail Always - 0
12 Power_Cycle_Count 0x0032 100 100 000 Old_age Always - 244
192 Power-Off_Retract_Count 0x0032 100 100 050 Old_age Always - 820
193 Load_Cycle_Count 0x0012 100 100 050 Old_age Always - 820
194 Temperature_Celsius 0x0002 171 171 000 Old_age Always - 32 (Lifetime Min/Max 6/47)
196 Reallocated_Event_Count 0x0032 100 100 000 Old_age Always - 6
197 Current_Pending_Sector 0x0022 100 100 000 Old_age Always - 0
198 Offline_Uncorrectable 0x0008 100 100 000 Old_age Offline - 0
199 UDMA_CRC_Error_Count 0x000a 200 200 000 Old_age Always - 5
|
A differentce of a several magnitudes :/
Of course I will still test it just In case, but I don't believe it likely ( then again, the odd problem which occurs with libAta, continues happening after device removal... so...er... )
So, I might go with the 3114 Controller ( more sockets, good price ) which I assume is the one vote you counted for =P
And if that controllers crap, ill just sell it to somebody else, no big loss ( well, apart from having lots of computer downtime )
it was really frustrating tho, I even tried using windows to get around the problem.
Went to seagates site and tried they're online disk analysis (via SMART ) , which they claim worked in netscape .. but as firefox counts for netscape imo, it doesn't work in netscape ( probably something to do with them using an ActiveX component .. retards.... ). So i loaded the evilness up in evilexplorer, where It tells me "your drive doesn't support SMART " .. but it obviously does.... linux can do it FINE! .. so I downloaded their little ISO for doing "dos" SMART analysis.. it doesn't even boot( oddly, i got it to work in QEMU, but it was useless there ) . So I here-forth dont trust seagate. |
|
Back to top |
|
|
mieses Tux's lil' helper
Joined: 28 Feb 2004 Posts: 110
|
Posted: Mon Jul 03, 2006 2:54 am Post subject: |
|
|
3ware has always provided solid linux support. I've had no problems with them. I think they make a 2 port sata controller.
<edit>
The gentoo docs once contained useful information dispelling some myths about "hardware" raid controllers.
This excerpt is copied from google cache:
Quote: | 2.b. SATA/PATA Raid Controllers
Controllers, Status and Types
Here is the status of the common SATA/PATA Raid Controllers:
Code: | Manufacturer Model Raid Type Status
VIA 8237 Software WORKING
Promise PDC202xx/PDC203xx Software WORKING
Silicon Image 3112[a],3512,3114 Software WORKING
Promise SX4000 (PDC20621) Hardware WORKING
3Ware Escalade 7xxx and 8xxx Hardware WORKING |
Difference Between Hardware And Software RAID
A Hardware raid controller is always an add-on card, it never comes distributed on a motherboard. It has a bios which you can enter before booting into any OS and usually supports 0,1,1+0,and 5 at a minimum. It has a full CPU onboard that does all raid calculations and I/O, and displays itself to the OS as configured by the raid controller (i.e. if you configure a single raid 5, from 3 drives, it will show up by the OS as one big drive). A Hardware RAID will always be faster than a software raid, and consumes MUCH less CPU time. A hardware RAID controller can come optionally with DIMM slots for caching, and possibly a battery backup for that cache. Hardware raid also limits the possible complexity of a OS driver because the raid functionality is performed exclusively in hardware.
A Software raid controller can be found in both add-on cards, and is distributed on many motherboards. A software controller may or may not have a BIOS, but the actual raid functionality is actually implemented by the driver in the OS. For this reason, you will NEVER find a software raid controller that can support a bootable RAID5. The OS will be able to see each drive as a standard hard drive, as it is not masked/transformed by the controller in any way. On 2.4 kernels, there was a module that could read many of the SATA controller's BIOSes, set up a linux software raid as specified by that bios, and create a psuedo device was accessible just like a hardware raid would be presented. This 'ataraid' module has not been ported to 2.6, and the 2.4 version does not support SATA controllers, only old PATA software raid controllers. To put it bluntly, a Software RAID controller is nothing more than a standard SATA/PATA controller with possibly a bios to store configuration information, this makes them extremely cheap to manufacture and is why you see them included on many motherboards. |
|
|
Back to top |
|
|
theJackalnz n00b
Joined: 05 Oct 2005 Posts: 32
|
Posted: Mon Jul 03, 2006 6:31 am Post subject: Thanks |
|
|
Thanks for that infos, most usefull. I've been reccomended by a friend just to ignore any of the card-supported "raid" functions as the linux kernel can emulate this faster than they can, and can use all that free ram for disk-caching as well ( with the only bad point being $OTHER_OS wont see it )
Unless of course there is a way to set up a software raid in linux that will be accessible in windows if I REALLY need to use it? |
|
Back to top |
|
|
NeddySeagoon Administrator
Joined: 05 Jul 2003 Posts: 54578 Location: 56N 3W
|
Posted: Mon Jul 03, 2006 10:15 am Post subject: |
|
|
theJackalnz,
If you really need Windows and Linux to share raid you need to use the BIOS (fake) raid.
Some are supported in the 2.6 kernel by dmraid, which replaces ataraid. _________________ Regards,
NeddySeagoon
Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail. |
|
Back to top |
|
|
|
|
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum
|
|