Revisions:
Originally published December 22 2006.

Summary:
The Dell CERC SATA 1.5/6ch RAID controller is a rebranded Adaptec card, available on Dell's PowerEdge 1800+ and NAS units.  Dell sold OEM solutions with this card, coupling it with 250GB or larger drives, touting a 1TB or larger storage solution in 4U.

The problem is, the card performs horribly, compared to other competing solutions.  This card is a poor representation of what SATA is capable of doing.  Avoid this card if you can.

Dell never officially made claims about the performance of this hardware, but providing it as an option, and refusing to publish benchmarks or performance data means that a lot of people have this card in production - and hate it.

If you need a SATA hardware RAID card for a file server, web server... get a 3Ware 9650SE card.

Details:
As Dell doesn't provide a BBU option with this card, I used this with the write cache disabled.  I had 3x320GB WD RE SATA 7200RPM drives in RAID5, and 2x120 WD SATA drives in RAID 1, attached.

I got around 40 MB/sec read and ~3 MB/s write with this card, on my RAID 5 set.  This is abysmal, considering its marketed as a "hardware RAID solution".  Updated the card firmware, as well as the server BIOS.  No use.

Contacting Dell was no help.  They scolded me for using a PE1800 with SATA disks; their SCSI solution was clearly the way to go.  At a premium of $2000 or more, it ought to be...

I made some changes in the BIOS, and discovered that if I disabled the onboard CERC2S SATA, and onboard LSI SCSI controller, I was able to get much better performance - around 40 MB read/write.

Characterizing the Poor performance:
I never spent the time to run benchmarks with this card.  Although I may have a chance in the future, now that I've FINALLY been able to get this card out of service.

However, I have rebuilt the RAID5 set on this card numerous times;

3xWD3200SD (RE, 8 MB cache, 7200RPM, SATA 1.5 Gbps, no NCQ), stripe size 64 kb.  Full rebuild with NO load on the drives took 82 hours.  EIGHTY-TWO HOURS.

Frustrated with the poor performance, I then added a 4th disk, so I had 4xWD3200SD plugged into a 3Ware 9650SE card.  Same 64 kb stipe.
With load on the disks, a rebuild to 42% completion took 3.23 hours.  The full rebuild around 7.5 hours.

Comparing just rebuild times, difference is night and day.  If I ever had a drive fail on the Dell CERC, I'd likely have another drive fail before the array even finished rebuilding.

I was scolded for having purchased a CERC 1.5/6ch SATA array for such a purpose.  The ONLY specifications that Dell had published on the array that they were able to give me over the phone was, "It uses an Intel processor for hardware RAID."

Conclusion:
I have a SATA "RAID" card that performs horribly.  While this card is perhaps a nice way to add 6 SATA ports to a system, the cost ($349 with cost of system) is quite hefty.  The 3Ware 9650SE 8 port comes in a $591 retail, but the performance difference, driver support, and management utilities are lightyears ahead.


 
Categories: 3Ware | IT

August 18, 2007
@ 04:34 PM
Categories:

August 10, 2007
@ 12:15 AM

Summary:

I recently got hold of an HP Media Center PC - a1720n.  There is so much junk in the preload.  If I bought this, and this was my first taste of Vista, I would return it and buy a Macintosh.  No wonder people hate Vista.

Background:

This a1720n Media Center PC is running Windows Vista Home Premium (OEM).  It is a Core 2 Duo 1.86 GHz with 1 GB DDR2 RAM, a 320 GB HDD, DVD-DL, Intel 945 onboard graphics, 10/100 ethernet, media card reader, 56k modem, etc.  4 SATA ports.  A decent system.

Comments:

Toolbars are coming out of the system.  I! opened! the! case! and! discovered! a! Yahoo! toolbar! in! the! CPU! cooling! fan!  Norton wants to activate a 60 day trial.  HP welcome center wants me to make a backup DVD set of my computer; I placed a blank in the drive, waited for it compile files.  The burn failed at 2% on Verbatim media when the application crashed.  I'm told to sign up for a high speed internet connection since I decided not to plug this infested machine into my LAN right away.  A pre-installed DVD MPEG2 decoder breaks the Windows built-in decoder; I can't even watch a DVD.  I don't want to Get Vonage, and I don't want to Sign Up for AOL.  Yes, I know my HP warranty is active.  No, I don't want to make a backup of my computer, and no I don't want to try 100's of games.  I don't have a printer installed... and no I don't want to buy one now.  I'm running low on disk space!  Apparently, the recovery partition (which is tiny partition) only has 14% free!  Didn't this computer come with 320 gigabytes or megabytes or whatever that stuff is called?

Icons are everywhere, flashing buttons, trial software... The desktop looks like a teenager's face with a serious acne problem.

If this is the way people - HP, in this case - sell PCs, people will continue to hate Bill Gates.  No, I don't see a sudden marketshare drop; most people have accepted that their computers will not function the way they want... and these are the people who are too lazy to do anything about it.  But for those who have had enough... they will switch to company that sells machines that work.  Out of the box.  And since this will be chalked up to Bill Gates' fault, that means that Steve has a new "convert".

The inherant problem with Windows is not the operating system itself.  With the release of Vista, it's not even the applications, viruses, or the hardware/driver problems.  It's the OEMs that are paid to install junk on the system, ruining the OOBE (out of box experience) that Microsoft worked so hard to develop.  They've stomped all over the fit and polish.

People who try other platforms aren't "converts" - they've just had enough of junk in the preload.

Conclusion:

I'm placing my Vista media in the drive now, and formatting the machine with a fresh, clean copy of Windows Vista Home Premium, using the license code on the side of the machine.

If you own this model computer, I suggest you do the same.  If you don't own a copy of the clean Vista install DVD, don't ask the people at the Genius Bar.  They won't be much help, I'm afraid.


 
Categories: IT | Media Center | Windows Vista

August 9, 2007
@ 09:21 PM

I've been looking at the Lenovo X61 tablet recently.  No, it's not for me... =)  But I will get to play with it!

Reviews seem to be hard to come by; perhaps it is due to it's recent release, the huge choice between models, and the inability of the general public to publish anything.  Most of what I've found has been related to the X61 or X61s.  While the X61 tablet is based on those machines, its particular configuration that makes it a tablet PC makes it unique.

Reviews I've found:

Platform details:

My particular hardware specifications: (via http://www-307.ibm.com/pc/support/site.wss/quickPath.do?quickPathEntry=7767-A4U&quickPathEntry.x=11&quickPathEntry.y=9)

  • X61 tablet, part 7767-A4U (it grinds my gears when people post specs without the part number when writing reviews)
  • L7500(1.6GHz)
  • 1GB RAM (will be upgraded!)
  • 160GB 5400rpm HD
  • 12.1in 1400x1050 LCD
  • Intel X3100
  • Intel 802.11agn
  • Bluetooth and Modem
  • 1Gb Ethernet
  • X6 UltraBase+CDRW/DVDRW
  • Secure chip, Fingerprint reader, 8c Li-Ion
  • Windows Vista Business (32-bit)

Once I get some time to play with one, I'll post some more details.


 
Categories: IT | ThinkPad

T. Haist and W. Osten, "An Optical Solution For The Traveling Salesman Problem," Opt. Express 15, 10473-10482 (2007)
http://www.opticsinfobase.org/abstract.cfm?URI=oe-15-16-10473

We introduce an optical method based on white light interferometry in order to solve the well-known NP–complete traveling salesman problem. To our knowledge it is the first time that a method for the reduction of non–polynomial time to quadratic time has been proposed. We will show that this achievement is limited by the number of available photons for solving the problem. It will turn out that this number of photons is proportional to NN for a traveling salesman problem with N cities and that for large numbers of cities the method in practice therefore is limited by the signal–to–noise ratio. The proposed method is meant purely as a gedankenexperiment.

http://www.opticsexpress.org/abstract.cfm?id=140598
http://www.opticsexpress.org/viewmedia.cfm?id=140598&seq=0

4D3DE0BF-BDB9-137E-C61F7482EABBDC28_140598.pdf (174.49 KB)
 
Categories: