Last night, I got the model number off of the cluster nodes and tried to use that too look up the specifications, but it didn't get me anywhere, so I had to pull the nodes themselves.
By opening the case, I found that each node has 512MB of ram - which is enough - and no PCI slot, making me wonder where I got the notion for using a low-profile PCI wireless card from.
After getting into the BIOS, I was able to find that the processor is a Pentium 4 2.4GHz Northwood core (533MHz bus, 512KB L2 cache), but unable to figure out if it has HT or not.
One of the CMOS batteries for the nodes died, so I decided to check the voltages - the dead one was jumping between 0.022 and 0.000, while the other two read 2.78 and 2.8 - and tossed the dead battery.
I then booted back into BIOS, and the node with the 2.78v CMOS battery refused to POST, and after some extensive digging around on the internet, I went and looked at the capacitors inside the node. Sure enough, one of them was bulging. Of all the times it could've vented, it decided to do it right after checking the CMOS battery. I transferred the CMOS battery and CD-ROM drive into the other node (the one that had the dead battery), and gave the "dead" node the floppy drive; I gave the stick of RAM to the node with the 2.8v battery. I was hoping I had a 16-volt 680µF capacitor in my collection of salvaged capacitors, but when I checked this morning, I certainly didn't.
Looking online, it seems like I technically can substitute with a capacitor with higher capacitance (µF), however since computers are sensitive components, it may cause a longer boot time or some sort of instability - I'll definitely pass on the notion in that case.
I also found that each node has the ability to boot from USB, which was surprising and nice, seeing that I'm not limited to using a 2.5-inch IDE hard drive for installation. It took a bit of digging around on the internet, but I eventually gave up and poked through the BIOS option very thoroughly, finding the USB option under the hard drive order; I test booted into the Linux stick and it worked just fine.
This morning, I grabbed Inquisitor and looked through my capacitor collection before plugging a node in to boot Inquisitor. Since I had more time, I figured out how to use Inquisitor properly for my intentions (hardware probing), and found out that the ethernet port is actually gigabit! I was quite surprised, but as I thought about it, it made sense since (1) it was built to be a business machine, and (2) the switch that connected the three nodes (when they were being used at my old work) was upgraded to a gigabit switch.
I didn't find out anything more about the processor though, which sucks, so I'll probably have to get the processor from the "dead" node to get the actual specifics.
I was thinking to maybe make one node as a summer computer, as it probably puts out less heat compared to any of the other desktops, but the problem is the lack of a hard drive. The options are to either use the 30GB hard drive from Sae-chan and to keep the drive usage light, or to run it as a fully live environment. I suppose using the hard drive wouldn't be that bad, but the lack of the drive cover is a bit unnerving.
My main point is to find out the exact processor in one way or another before I go digging around for a Linux distribution to test with - last thing I need is to have a kernel limitation like I did with the Half-top.
No comments:
Post a Comment