Son of a Son of a SysAdmin
Monday, October 3, 2011
Wanted: Broken Garmin Nuvi 295w
Just looking for a Broken Garmin Nuvi 295w that I can pull apart to see how it looks. I am looking at hacking on the code Garmin provided, but don't want to take the chance on breaking my working Garmin 295w.
Thursday, January 13, 2011
Foxconn Small Form Factor R10-D2 WinXP experience
So one of the online retailers had a SFF Foxconn (R10-D2) box on sale in December with 4GB of ram and I added a sata DVD-RW into the mix for another $18. Not bad for $145 shipped.
A dual core Atom 1.6Ghz with Hyperthreading. I wasn't expecting much, but it had to better than the Dell Celeron my mom's been complaining about. Get around to it, enable AHCI (native SATA) and install Ubuntu-10.10 on it. Wow. This is not bad.
Let's try Windows XP, since that's what the Dell is running. No Go. Course, I am a bit of a Windows idiot, having done Solaris for nearly 20 years. Try a couple of times, No Go. Eventually email Foxconn support, and they say they've installed WinXP on that box. I then realize that I've enabled AHCI support in the bios, and it dawns on me that the drivers probably aren't found, especially for Windows XP.
It doesn't help that Foxconn didn't put the Intel Raid Drivers up in the driver area for the R10-D2, or even indicate what kind of motherboard was in the R10-D2 unit.
Eventually go back to IDE emulation for SATA and WinXP installs. Hmm. Email Foxconn support again. Finally, they point me at the Intel Raid Drivers up in the D51S motherboard section.
I use something called nlite to drop all the drivers into a slipstream'd WinXPSP3 disk and off I went. Last thing I needed was my USB floppy, with the 32-bit drivers extracted from the Intel Raid driver disk (the Foxconn Motherboard is called the D51S) in the top level directory with the magic oemsetup.txt file, and WinXPSP3 is happily installing. This was worth the effort and frustration, to learn that the Intel Raid Drivers is where they hide the native SATA (AHCI) drivers for WindowsXP, and that I no longer have to install in IDE mode, then do some very convoluted procedure to get AHCI working after an IDE install.
A dual core Atom 1.6Ghz with Hyperthreading. I wasn't expecting much, but it had to better than the Dell Celeron my mom's been complaining about. Get around to it, enable AHCI (native SATA) and install Ubuntu-10.10 on it. Wow. This is not bad.
Let's try Windows XP, since that's what the Dell is running. No Go. Course, I am a bit of a Windows idiot, having done Solaris for nearly 20 years. Try a couple of times, No Go. Eventually email Foxconn support, and they say they've installed WinXP on that box. I then realize that I've enabled AHCI support in the bios, and it dawns on me that the drivers probably aren't found, especially for Windows XP.
It doesn't help that Foxconn didn't put the Intel Raid Drivers up in the driver area for the R10-D2, or even indicate what kind of motherboard was in the R10-D2 unit.
Eventually go back to IDE emulation for SATA and WinXP installs. Hmm. Email Foxconn support again. Finally, they point me at the Intel Raid Drivers up in the D51S motherboard section.
I use something called nlite to drop all the drivers into a slipstream'd WinXPSP3 disk and off I went. Last thing I needed was my USB floppy, with the 32-bit drivers extracted from the Intel Raid driver disk (the Foxconn Motherboard is called the D51S) in the top level directory with the magic oemsetup.txt file, and WinXPSP3 is happily installing. This was worth the effort and frustration, to learn that the Intel Raid Drivers is where they hide the native SATA (AHCI) drivers for WindowsXP, and that I no longer have to install in IDE mode, then do some very convoluted procedure to get AHCI working after an IDE install.
Monday, January 10, 2011
4 systems, built same way, patch outcome different
Built a clustered system, 4 nodes, with Solaris 10 months ago. Finally hit enough problems that it's decided that finally we get our patch upgrade. zfs root rocks again, lucreate, lumount, pca, luactivate, init 6 and we're on the patched OS.
1) system 1 - /usr/java still pointing at jdk/jdk_1.6.0_14, appearently patch install failed. logs disagree... Reinstall patches and fix /usr/java *again*.
2) system 2 - /usr/java now pointing at jdk/jdk_1.5.0_27, huh? 1.5 now? Fix link.
3) system 3 - /usr/java now pointing at jdk/jdk_1.6.0_23, expected
4) system 4 - /usr/java now pointing at jdk/jdk_1.5.0_27. sheesh
You'd expect that 4 systems, built exactly the same way, patched exactly the same way, would have the same outcome from *deterministic* code from the patch install. The law on unintended consequences strikes again.
1) system 1 - /usr/java still pointing at jdk/jdk_1.6.0_14, appearently patch install failed. logs disagree... Reinstall patches and fix /usr/java *again*.
2) system 2 - /usr/java now pointing at jdk/jdk_1.5.0_27, huh? 1.5 now? Fix link.
3) system 3 - /usr/java now pointing at jdk/jdk_1.6.0_23, expected
4) system 4 - /usr/java now pointing at jdk/jdk_1.5.0_27. sheesh
You'd expect that 4 systems, built exactly the same way, patched exactly the same way, would have the same outcome from *deterministic* code from the patch install. The law on unintended consequences strikes again.
Wednesday, September 22, 2010
Autonegotiation doesn't
Standing up a Snorcle x4540 storage server. Nice Cisco 4849 switches, and Solaris 10U9. But wait, the stupid NVidia NGE driver autonegotiates 100/Half to the Gigabit Network switch. Whiskey-Tango-Foxtrot. My goodness. Years and years of hardware development and standards and two name brand devices can't figure it out. Great.
Start pounding on the various incarnations of the adv_*_cap flags with ndd to see if I can get a clue why it's doing what it's doing. Big chart brewing, and I'm probably gonna have to talk to some guy in a foreign language to see if this problem has a known resolution, or if it's gonna take hammering it for a couple of hours to find the 2^12 combinations that gets me 1000Mbs. At least the network guy is a charm to work with so I know we'll figure it out the hard way.
Start pounding on the various incarnations of the adv_*_cap flags with ndd to see if I can get a clue why it's doing what it's doing. Big chart brewing, and I'm probably gonna have to talk to some guy in a foreign language to see if this problem has a known resolution, or if it's gonna take hammering it for a couple of hours to find the 2^12 combinations that gets me 1000Mbs. At least the network guy is a charm to work with so I know we'll figure it out the hard way.
Friday, September 17, 2010
Office Moves
Another huge time waster. Not only do I get a smaller cube, but right across from the Group Director's office who runs the "can do no wrong" NOC. I'm so happy. NOT.
Tuesday, September 14, 2010
How dumb do you think we are?
I work in a support organization who has to call support organizations. Often times, I've already done 75% of the "debugging" before I've called so I can effectively bypass that. But one of my vendors has changed, and now they're relying on automated tools, and associated "tools" who make annoying emails about the status of our systems without one clue about what they talk.
An example, I built out a pair of systems about 6 weeks ago. Having some storage problems, and send of the appropriate diagnostics. The rocket scientist on the other end of the line tells me I have a bad disk and I need to replace it. Oh, yeah, Which disk in my SAN array are you talking about? "Oh". mumbles for a while, says he'll get back to me. Things aren't quite right and said rocket scientist again says I need to update the patches to fix things up. I asked him for the specific bug id's he's referring to, along with the patches. "Oh." mumbles for a while. "But our backline says you're down rev'd on patches". Me - Really. 6 weeks ago, this was completedly patched and you've not yet identified any of the problems I've pointed out to you with appropriate bug ids. Yeah, you're using a tool that says patches are out of date. I got one of those too. I don't need you to tell me that. I need you to tell me "here is the bug, here is the bug id and here is the patch or temporary patch". Don't be telling me I need to upgrade my patches on your whim and your stupid tool just because you are out of ideas. What, do you think every system is in a lab and can be upgraded willy-nilly on your say so?
Please buy a clue. and if not, go get a job where they ask "Do you want fries with that?"
An example, I built out a pair of systems about 6 weeks ago. Having some storage problems, and send of the appropriate diagnostics. The rocket scientist on the other end of the line tells me I have a bad disk and I need to replace it. Oh, yeah, Which disk in my SAN array are you talking about? "Oh". mumbles for a while, says he'll get back to me. Things aren't quite right and said rocket scientist again says I need to update the patches to fix things up. I asked him for the specific bug id's he's referring to, along with the patches. "Oh." mumbles for a while. "But our backline says you're down rev'd on patches". Me - Really. 6 weeks ago, this was completedly patched and you've not yet identified any of the problems I've pointed out to you with appropriate bug ids. Yeah, you're using a tool that says patches are out of date. I got one of those too. I don't need you to tell me that. I need you to tell me "here is the bug, here is the bug id and here is the patch or temporary patch". Don't be telling me I need to upgrade my patches on your whim and your stupid tool just because you are out of ideas. What, do you think every system is in a lab and can be upgraded willy-nilly on your say so?
Please buy a clue. and if not, go get a job where they ask "Do you want fries with that?"
Thursday, August 26, 2010
you want what version?
As most companies do, mine uses co-lo's to deal with the power, cooling, network, storage and other logistics. Recently, we needed some storage and the co-lo said that we could only run the qlogic's driver from Solaris 10U5. I was like Whiskey-Tango-Foxtrot? Are you kidding me? You haven't updated your support matrix in 3 years? I was blown away.
Fortunately, after a fair amount of hang wringing and arguing, we finally got them to agree to a more recent version, and only required a backout of a single patch. I do love Live Upgrade with ZFS root. create BE, remove patch, run pca (Patch Check Advanced) against the ABE to verify that the right patch version will be activated, activate the ABE and reboot. Firedrill averted.
Fortunately, after a fair amount of hang wringing and arguing, we finally got them to agree to a more recent version, and only required a backout of a single patch. I do love Live Upgrade with ZFS root. create BE, remove patch, run pca (Patch Check Advanced) against the ABE to verify that the right patch version will be activated, activate the ABE and reboot. Firedrill averted.
Subscribe to:
Posts (Atom)