Archive for the ‘hardware’ Category

Newer Entries »

Linux (ubuntu Karmic Koala) on HP Elitebook 2530p

Saturday, October 24th, 2009

Finally got my new laptop (the old one served me well), an HP elitebook 2530p. I had my doubts on the ability to be able to get suppport for every piece of hardware in this laptop under Linux, but it was doable. Not without issues, but still fairly easy.

I opted for the latest ubuntu, Karmic Koala (9.10), which on the time of writting was release candidate. The installation went smoothly. The windows partition got resized without issues and after a few clicks on the next button the system was up and running. With encrypted ext4 filesystem and all.

Working without a hitch:

  • acpi (suspend, fan control, speedstep)
  • sound
  • graphic
  • wired network

Working with some debugging

  • Initially the wireless would not work with my wrt54g. It worked with other wifi hotspots. Symptoms were that the laptop could not a dhcplease from the router. Tried to flash the router with dd-wrt firmware instead of the tomato firmware. Did not help. Breakthrough came when I reset the router settings to “Default”. Then it worked with dd-wrt. I reflashed it with tomato and it still worked. So it is possible to enable “some setting” in the wrt54g router that cause problems with this laptop. My other laptops did not inhibit these issues. But then again. The laptop can not really be blamed.
  • My external Huawai E180 did not work. After some googling, it turned out to be a bug introduced in a specific kernel release.  The solution was to downgrade the kernel to a previous version. The only one available was 2.6.31-9-rt, which I choose and that works just fine.
  • The built-in Qualcomm Gobi 1000 3G module requires a firmware and a firmware loader to function. The firmware loader can be found here. It is just a little .c program and a set of udev rules. After that you will need the firmware. That can be obtained from your windows partition. If you do not have the driver for the windows environment it can be obtained from hp. The Gobi chipset is a fullblown arm cpu which can function as a 3G modem for various types of 3G nets. Different firmwares allows access to different networks. This piece of information gives an overiew. For my part I needed the 6. th version of the firmware as I live in Europe/Denmark.

All in all a couple of minor issues. I am really satisfied with this latop and the way Linux works on it. Will hopefully be a good laptop for me in the next couple of years.

Posted in hardware, unix, work | No Comments »

New server again again – based upon an Intel D945gclf2 – dual core Atom 330

Sunday, May 24th, 2009

Well. More time used on my server. More downtime. More annoyance towards my wife (who have come to be dependent on the mailsetup).

My previous server simply consumed too much power. I basically just plugged it in and it ran. It is a nice little board for around 500dkk:

edison% cat /proc/cpuinfo
processor    : 0
vendor_id    : GenuineIntel
cpu family    : 6
model        : 28
model name    : Intel(R) Atom(TM) CPU  330   @ 1.60GHz
stepping    : 2
cpu MHz        : 1595.997
cache size    : 512 KB
physical id    : 0
siblings    : 4
core id        : 0
cpu cores    : 2
apicid        : 0
initial apicid    : 0
fdiv_bug    : no
hlt_bug        : no
f00f_bug    : no
coma_bug    : no
fpu        : yes
fpu_exception    : yes
cpuid level    : 10
wp        : yes
flags        : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe nx lm constant_tsc arch_perfmon pebs bts pni monitor ds_cpl tm2 ssse3 cx16 xtpr lahf_lm
bogomips    : 3191.99
clflush size    : 64
power management:

processor    : 1
vendor_id    : GenuineIntel
cpu family    : 6
model        : 28
model name    : Intel(R) Atom(TM) CPU  330   @ 1.60GHz
stepping    : 2
cpu MHz        : 1595.997
cache size    : 512 KB
physical id    : 0
siblings    : 4
core id        : 1
cpu cores    : 2
apicid        : 2
initial apicid    : 2
fdiv_bug    : no
hlt_bug        : no
f00f_bug    : no
coma_bug    : no
fpu        : yes
fpu_exception    : yes
cpuid level    : 10
wp        : yes
flags        : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe nx lm constant_tsc arch_perfmon pebs bts pni monitor ds_cpl tm2 ssse3 cx16 xtpr lahf_lm
bogomips    : 3191.98
clflush size    : 64
power management:

processor    : 2
vendor_id    : GenuineIntel
cpu family    : 6
model        : 28
model name    : Intel(R) Atom(TM) CPU  330   @ 1.60GHz
stepping    : 2
cpu MHz        : 1595.997
cache size    : 512 KB
physical id    : 0
siblings    : 4
core id        : 0
cpu cores    : 2
apicid        : 1
initial apicid    : 1
fdiv_bug    : no
hlt_bug        : no
f00f_bug    : no
coma_bug    : no
fpu        : yes
fpu_exception    : yes
cpuid level    : 10
wp        : yes
flags        : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe nx lm constant_tsc arch_perfmon pebs bts pni monitor ds_cpl tm2 ssse3 cx16 xtpr lahf_lm
bogomips    : 3127.96
clflush size    : 64
power management:

processor    : 3
vendor_id    : GenuineIntel
cpu family    : 6
model        : 28
model name    : Intel(R) Atom(TM) CPU  330   @ 1.60GHz
stepping    : 2
cpu MHz        : 1595.997
cache size    : 512 KB
physical id    : 0
siblings    : 4
core id        : 1
cpu cores    : 2
apicid        : 3
initial apicid    : 3
fdiv_bug    : no
hlt_bug        : no
f00f_bug    : no
coma_bug    : no
fpu        : yes
fpu_exception    : yes
cpuid level    : 10
wp        : yes
flags        : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe nx lm constant_tsc arch_perfmon pebs bts pni monitor ds_cpl tm2 ssse3 cx16 xtpr lahf_lm
bogomips    : 3191.95
clflush size    : 64
power management

Posted in hardware | No Comments »

New server ….

Monday, April 13th, 2009

… well, almost. To let one of my boys use the nVidia 7900 GT graphics card I bought for them, I had to move some hardware around. My Aopen i915gmm-hfs motherboard was taken out of my server and put into Benjamins computer. Instead my server got his ASUS A8N-VM CSM motherboard. To keep power consumption of my server down, I decided to enable cool’n’quiet, use a 2.5″ diskdrive for the root partition and use powertop to minimize the number of interrupts. Tomorrow I have to get rid of the built in nvidia NIC (it generates 100 ints/sec) and use an intel pro100 or pro1000

In any case this is only temporary, until I can get an atom or arm based mainboard with 4 sata connectors without having to shell out millions of dollars on it :-)

Ohh yeah, also “upgraded” from Debian to Ubuntu. Ran into some anoyances with regards to /var/run and /var/lock combined with a seperate /var partition. Ubuntu is not really server quality yet if you can not move /var to a seperate partition without problems. Sorry, but it really is not!

Posted in hardware, unix | No Comments »

Debian Linux on a pSeries 610 (7028 – 6C1)

Friday, August 15th, 2008

After a couple of tries I got debian installed on an IBM pSeries 610 (7028-6C1). I got the machine through my work and it is also hosted there. I will write a short descripton on how to do it and more important what pitfalls to avoid. The server is named after a famous inventor. The name is a spin of the fact that I work as a Unix Administrator in a large Telco Company in denmark.

Enough of me talking. Some juicy details:

zensonic@bell:~$ uname -a
Linux bell 2.6.18-6-powerpc64 #1 SMP Wed Jun 18 06:03:18 UTC 2008 ppc64 GNU/Linux

zensonic@bell:~$ cat /proc/cpuinfo
processor : 0
cpu : POWER3 (630+)
clock : 375.000000MHz
revision : 1.4 (pvr 0041 0104)

processor : 1
cpu : POWER3 (630+)
clock : 375.000000MHz
revision : 1.4 (pvr 0041 0104)

timebase : 93746809
platform : pSeries
machine : CHRP IBM,7028-6C1

zensonic@bell:~$ top | head -5
top – 12:26:38 up 3:00, 1 user, load average: 0.00, 0.00, 0.00
Tasks: 35 total, 1 running, 34 sleeping, 0 stopped, 0 zombie
Cpu(s): 0.1%us, 0.2%sy, 0.0%ni, 99.6%id, 0.1%wa, 0.1%hi, 0.1%si, 0.0%st
Mem: 8108636k total, 51796k used, 8056840k free, 6712k buffers
Swap: 1502036k total, 0k used, 1502036k free, 9372k cached

zensonic@bell:~$ dmesg | grep -i bogo
Calibrating delay loop… 186.88 BogoMIPS (lpj=373760)

zensonic@bell:~$ more /proc/rtas/sensors
RTAS (RunTime Abstraction Services) Sensor Information
Sensor Value Condition Location
********************************************************
Key switch: Normal (read ok) —
Power source: AC (read ok) —
EPOW Sensor: EPOW Reset (read ok) —
Temp. (C/F): 29 / 84 (normal) Planar #1
Surveillance: 0 (read ok) —
Fan (rpm): 2070 (normal) Fan #1
Fan (rpm): 2820 (normal) Fan #2
Fan (rpm): 3210 (normal) Fan #3
Fan (rpm): 3000 (normal) Fan #4
Voltage (mv): 5129 (normal) Planar #1
Voltage (mv): 3129 (normal) Planar #1
Voltage (mv): 5161 (normal) Planar #1
Voltage (mv): 12077 (normal) Planar #1
Powersupply: 3 (normal) Planar #3 at Voltage #1
Powersupply: 3 (normal) Planar #3 at Voltage #2
Powersupply: 0 (non existent) Planar #3 at Voltage #3

Posted in hardware, unix | No Comments »

Enabling 32MB memory on the linksys WRT54G router (version 2.2 XH only!)

Friday, July 13th, 2007

Once in a while luck strikes. I was really lucky buying an wrt54g when they first appeared. I got a revision 2.2 which meant that they had ironed out the bugs from the first revisions, but not yet decided to cripple the unit as they did in revision 5 and 6. Furthermore I was lucky enough to get a revision 2.2XH on which linksys, for unknown reasons, has put 32MB of memory (normally 16MB for other revisions). Linksys has disabled half of the 32MB to make the unit look like every other wrt54g. Here is how to enable it.

Firstly, no cheating:

~ # uname -a
Linux dd-wrt 2.4.34-pre2 #174 Fri Sep 15 20:38:23 CEST 2006 mips unknown

How much memory to begin with:

~ # cat /proc/meminfo | head
total: used: free: shared: buffers: cached:
Mem: 14516224 13836288 679936 0 462848 5664768
Swap: 0 0 0
MemTotal: 14176 kB

Perform some magic:

~ # nvram set sdram_init=0×008
~ # nvram set sdram_ncdl=0×000
~ # nvram commit
nvram_commit(): end
~ # reboot

How much do we have now?

~ # cat /proc/meminfo | head
total: used: free: shared: buffers: cached:
Mem: 31113216 18386944 12726272 0 1835008 8638464
Swap: 0 0 0
MemTotal: 30384 kB

Ofcourse I use a custom firmware on the unit, but more on that in another post.

Posted in hardware | No Comments »

Asus A8N CSM … Lovely little mainboard

Sunday, February 25th, 2007

After having sold my graphic card due to a descission of not playing battlefield anymore I had little use for my athlon 64 board with an AGP socket. Actually the board was useless as I didn’t even have an old AGP card laying around. Instead of buying an already obsolete AGP graphics card I invested some money in a socket 939 PCI Express Board.

The obvious choice would have been to buy an Nvidia Nforce4 based board. That would however, have required me to buy either an expensive graphic card (tempting but would have been a personal failure as I would end up playing BF2 again).

Instead I choose the Asus A8N CSM board. Based around an Nvidia 6150 northbridge and a Nvidia 430 southbridge this board is worth every penny. Onboard everything and decent performance for $60!! The review that convinced me.

So after installing the board and playing around a bit with it I consider it my best mainboard purchase in quite a while. I was really interessted in the picture and sound quality and I have to say that I’m impressed. It is a wonder that the Analog Devices chips are not used more than they are. I have a hard time telling the difference between my Audigy 2ZS and the onboard AD1986A. Reviews has stated that this was to be expected but I thought they overhyped the chip, but this isn’t the case.

I will give it 8 out of 10. Deducting 1 point for Asus not shipping the tv-out bracket with the board and 1 point for not enabling over/underclocking and voltage in the bios.

Posted in hardware | No Comments »

“Refurbished to Maxtors standards”

Sunday, December 10th, 2006
Maxtor MaxLine II+ Arrgghh. My refurbished Maxtor Maxline II+ drive died today! “Refurbished to Maxtors standards” they wrote on top of it! Truely a lovely set of standards. Not even a month it lasted. Luckily for me it only contained mythrecordings. Anyhow I’m done with Maxtor.

My replacement drive is going to be the Seagate Baracuda 7200.10 or the Segate Baracuda ES (7200.10 with enterprise firmware). Whatever I can get my hands on.

As evidence:

SMART Self-test log structure revision number 1
Num Test_Description Status Remaining LifeTime(hours) LBA_of_first_error
# 1 Short offline Completed: read failure 60% 484 12845440
# 2 Short offline Completed: read failure 60% 469 12845440
# 3 Extended offline Completed: read failure 40% 463 12845440
# 4 Short offline Completed: read failure 60% 462 251138448
# 5 Short offline Completed without error 00% 439 –
# 6 Short offline Completed without error 00% 417 –
# 7 Short offline Completed without error 00% 395 –
# 8 Short offline Completed without error 00% 372 –
# 9 Short offline Completed without error 00% 350 –
#10 Short offline Completed without error 00% 328 –
#11 Extended offline Completed without error 00% 308 –
#12 Short offline Completed without error 00% 305 –
#13 Short offline Completed without error 00% 283 –
#14 Short offline Completed without error 00% 261 –
#15 Short offline Completed without error 00% 238 –
#16 Short offline Completed without error 00% 216 –
#17 Short offline Completed without error 00% 194 –
#18 Short offline Completed without error 00% 171 –
#19 Extended offline Completed without error 00% 151 –
#20 Short offline Completed without error 00% 149 –
#21 Short offline Completed without error 00% 127 –

Posted in hardware | No Comments »

Server upgrade

Tuesday, November 21st, 2006

I finally did, what I should have done a long time ago. I upgraded my server. I added 256MB of memory so that it in total have 1GB. But more important I added a disk. After installing mythtv at home, my disk space is ever shrinking :-) And then I also took the time to redo my partitioning scheme using LVM for everything except / and /boot. The stats speak for them self

However I still can not justify doing raid at home as each disk consumes around 10W and I need at least 6 to make raid5 worthwhile IMHO. Instead  I have installed smartmon-tools and set it to run short tests every day at 24.00 execept saturday where it runs a long test. Hopefully this should be sufficient together with my backup scheme (which all people have, right!?)

Posted in hardware | No Comments »

Newer Entries »