SuperMicro SuperServer 6018R-MT / X10DRL-i

My company recently purchased a SuperMicro SuperServer 6018R-MT which contains a X10DRL-i motherboard.  The BMC in the server would not work after configuring the IP address, despite upgrading the BMC firmware.  The server came from the factory installed with version 1.50, and we updated to the latest version from the SuperMicro website (which was 1.64).  We originally thought that it could be related to the BMC now running on a 1 Gbps port instead of the 100 Mbps port speed of previous SuperMicro BMCs.  We tried different switch ports, switches, Ethernet cables, switch port settings, and lots of other things in an attempt to get it working.  After a few frustrating hours, we plugged a laptop directly into the BMC port, configured the Ethernet port of the laptop to an IP address in the same subnet as the BMC, and to our surprise we were able to get to the web interface!  It seems that the gateway address was not being sent from the BIOS to the BMC.  After manually setting the gateway address through the web interface, we were able to access the BMC with no issues.  Another alternative is to use the ipmicfg utility available from SuperMicro’s website.

I thought I’d post this in case anyone else runs into issues with this specific system/motherboard.  Hopefully it saves you some time!

South Carolina House Passes Tax Break for Data Centers

I just saw an article in The State about the South Carolina House passing tax incentives for new data center builds. Basically, it waives sales tax on power consumption, matching similar incentives in North Carolina and Virginia.

This should be interesting to see how South Carolina stacks up versus North Carolina for future builds from the likes of Google, Facebook, Apple, etc. South Carolina taxes are overall lower anyway (NC currently has the 6th highest gas tax in the country – versus 47th for SC), so it should give them a slight advantage. Additionally, the Upstate of South Carolina is very close to Atlanta, which is a major hub for Internet traffic.

Intel DG33TL – Supported CPUs

Despite what is on Intel’s website, the Intel DG33TL does not support Core 2 Quad Q9550 CPUs and has trouble with 1333 MHz FSB chips in general. I also ran into problems with Core 2 Quad Q6600 chips, but not on every board.

Most of the time, the board will post and will work somewhat, but things like virtualization support in the BIOS won’t show up. Other times, the board will not post and doesn’t seem to work.

This board also seems to be very picky about RAM. If you put some RAM in the first and third slots, it won’t post, but if you put it in the second and fourth slots (and put a different brand/model RAM in the first and third slots), it will post.

I’d steer clear of this board unless you are looking to use it with a 1066 MHz or less FSB CPU with either single core or dual cores.

Intel DG41WV Motherboard = fail

Well, after a long break, I’m going to start posting again. 🙂

We purchased some Intel DG41WV LGA775 Motherboards back about August 2011 because they supported DDR3 RAM, which is significantly cheaper than DDR2 RAM. We had a few systems that customers wanted to upgrade to 8 GB RAM, and the old motherboards only supported 4 GB.

We had some weird issues getting them up and running, but now that they are in service, most of them seem to run without issues. However, one of the boards is having major issues running XenServer 6/5.6, so we went to see if there was a new BIOS update.

SURPRISE – Intel and NewEgg both have lowered the maximum supported RAM to 4 GB. Wow. How can you change the maximum supported RAM after the fact?

This might spell the end for our purchase of Intel motherboards.

We recently purchased a number of Gigabyte GA-G41MT-S2PT and they work flawlessly. We’ll probably purchase some more and throw the Intel boards on eBay.

I’m very disappointed in Intel and NewEgg. We would not have purchased these boards if they stated that they only supported 4 GB when we initially purchased them.


The opinions here are solely my own and do not represent the opinions of any of my employers – past, present, or future. I wanted to post this disclaimer since there was a recent e-mail sent out that advised us to post this disclaimer if we have any blogs.

Linux Hardware Compatibility

I have yet to find a reliable source for learning whether a motherboard or other piece of hardware is going to work in Linux. There are some sites and forums out there that have bits and pieces, but I have not found a reliable, up-to-date, and complete source for the latest hardware. Especially since there are so many viable distros – each with their own level of compatibility.

If anyone knows of a site that I haven’t found yet, please shoot me an e-mail to my first name at


From the article:

Ext3 tops out at 32 tebibyte (TiB) file systems and 2 TiB files, but practical limits may be lower than this depending on your architecture and system settings—perhaps as low as 2 TiB file systems and 16 gibibyte (GiB) files. Ext4, by contrast, permits file systems of up to 1024 pebibyte (PiB), or 1 exbibyte (EiB), and files of up to 16 TiB. This may not be important (yet!) for the average desktop computer or server, but it is important to users with large disk arrays.


The single most dramatic improvement in ext4 is in file and file system size. Thus, the users who are most likely to need ext4 are those who have more than a few terabytes of disk space. The list of features in Table 1, though, may present some other tempting improvements. For instance, you might want to try ext4 if you have directories with huge numbers of subdirectories or if you need timestamps accurate to less than a second.

Very nice. I’m sure we’ll be concerned about 2 TiB files soon enough…

Kubuntu 7.10 (Gutsy Gibbon)

I know I’m a little late to the party in terms of installing Kubuntu 7.10. However, I finally installed it on my Dell 600m this weekend. It worked out of the box with absolutely no tinkering which was a pleasant surprise. Kubuntu is surprisingly polished and I’m pretty impressed so far.

After updating all of the packages and rebooting, I found myself with a non-functional wireless card. Having had this happen in the past, I knew where to look. The firmware was deleted from the system when the new kernel was installed and the firmware needed to be downloaded.

To get the wireless card working again, I did the following (procedure modified for use with Kubuntu from How to enable Wireless Lan on [wifi] FC-6):

1. Download ipw2200-fw-3.0.tgz from Intel® PRO/Wireless 2200BG Driver Firmware.

2. Unarchive ipw2200-fw-3.0.tgz.

3. Move all .fw files to /lib/firmware/2.6.22-14-386 (the exact directory may vary depending on your kernel version).

3. /sbin/modprobe -r ipw2200

4. /sbin/modprobe ipw2200

Your wireless card should be working again. Enjoy.

Insane Comment Spam

If anyone knows of a good comment spam killer, please let me know by e-mail. I am probably going to disable comments until I can figure out what to do about it. Akismet is not working anymore.

CentOS 5 and RHEL 5

CentOS 5 was finally released this week after RHEL 5 was released earlier this month. I did an install on a spare server to play around with it, and so far so good. I may post a more in depth review after I’ve logged some more time playing around.

I just read an article in eWeek titled Red Hat should lighten up.

I think Red Hat’s current policy of requiring subscriptions for access to RHEL is flawed. CentOS has a huge userbase which could contribute more directly to the development and community around RHEL if Red Hat opened up RHEL to users who don’t need support.

I understand that there are definite costs to having mirrors for all of the users. Many companies have stepped up and offered mirrors, and I think they would have no problem continuing to do so if Red Hat opened RHEL.

Since CentOS currently has a 2-3 day lead time to download the new source RPMS and compile them for the OS’s after they are released, why not do the same for the free version of RHEL? Add that as another advantage to buying a subscription — priority updates.

If there were a free version of RHEL, it could also change Oracle’s plans on Unbreakable Linux, which like CentOS, is just another rebadged RHEL.

I think that what bugs me the most is that there’s so much wasted productivity in the rebadging/recompiling/etc that could be better utilized in the further development of the OS.

I do understand Red Hat’s position on RHEL. I do see potential downsides to opening it up — the main one being that some users that now pay for subscriptions might switch to the free version. However, I think the potential benefits outweigh the risks.