Performance Tuning

Adding AHCI support to CentOS after it was already installed in IDE mode

Sometimes when you’re using a server, you’ll notice that the disk i/o is slower than it ought to be, or is using a lot more cpu than it should be during disk i/o. In some cases, this would be because the BIOS is configured to use your sata drives in legacy IDE mode instead of

How to set up source nat routing (SNAT) in iptables to load balance outbound connections across multiple IP addresses

In our earlier days, before IOFlood, some of us ran web based proxy services. These were popular at workplaces or schools for accessing websites that were blocked there, such as gmail and myspace (remember myspace?). One common problem that came up, was that with so many users each sharing one IP on one server, our

Making sense of Linux memory usage, Part 1: how to read “top” in CentOS.

We get this question a lot, about “Why is my Linux server using so much ram?”. In many cases, it really isn’t using much ram at all, but just to be sure we always have to check. At least half the time, really very little is being used, but the way that Linux reports ram

nf_conntrack: table full, dropping packet — A solution for CentOS Dedicated Servers

A common problem you may experience is sluggish performance or disconnections from your Centos dedicated server, even though there is sufficient CPU, ram, disk i/o, etc. After some troubleshooting, you may come to believe you are being DDoS attacked, but you don’t see an unusual amount of traffic, and there’s no single IP or handful

Troubleshooting high linecard CPU use / exhaustion on Brocade MLX and MLXe routers

Today we’re going to be talking about a poor design decision in the Brocade MLX and MLXe series routers that can lead to CPU exhaustion on your router line cards. Overall we’re very happy with the Brocade MLX. After all, every routing platform has its quirks, so one small problem that is easily solved is

CloudFlare agrees: Disk performance is improved when you get rid of hardware raid

I know this article is a little old, but I wanted to bring it up as it raises an interesting point. We’ve had a few articles here about hardware raid, detailing when we think it’s a good idea, and when it’s a waste of money (or worse). In general, if you’re looking for “big DD

Real world performance of Brocade Foundry FESX-448-PREM — CPU exhaustion

In case you’re still using the Brocade / Foundry FESX-448-PREM for doing routing on your network, I thought I would share some of the performance limitations we’ve seen on that platform. We’ve long since upgraded our network to use newer and more powerful Brocade MLX routers, but wanted to share our experience so others could

Update: Hardware vs Software RAID: The great debate.

Hi Everyone, If you recall our previous article Hardware raid or software raid? Our take on the great religious debate, we made the case that hardware raid served virtually no purpose, and software raid was better in all important ways. Although we still find most of the information in that article to be true, we are

Hardware raid or software raid? Our take on the great religious debate

I get asked this question a lot: “Do I need hardware raid, or is software raid good enough?” You may have noticed I/O FLOOD only offers software raid, so it will come as no surprise to you that we think software raid is “good enough”. What might surprise you however, is that we believe that