Archive
Posts from 2011
Kudzu removed from RHEL6
With the introduction of RHEL6 the kudzu hardware hardware manager was removed. All of the functionality that was once a par of kudzu has been integrated into the kernel and udev, as evidenced by this e-mail correspondence with one of Redhat's support engineers: "Kudzu is removed from rhel6. The kernel should be taking care of module loading from this point onwards. When it enumerates the device through its own methods or udev…
$ read more →Measuring hard drive seek times
I've been fascinated with I/O and file system performance for years, and chose prefetch.net as my domain name after reading about pre-fetching algorithms in the book UNIX file systems (a great book that I need to read again). Since most applications access data that is not laid out sequentially on a hard drive platter, seek times come into play when you start to look at getting the most bang for your buck on random I/O workloads. This past week I saw this first hand. I was playing around with some disk drives and started to wonder how many seeks I could get per second…
$ read more →Speeding up SSH (SCP) data transfers
I'll be the first to admit that I'm an SCP addict. It doesn't matter what kind of data I'm working with, if it can be turned into an object that I move around with scp I'm in! One thing I've always noticed with scp is the dismal out of the box performance. I read quite some time back on the openssh mailing list that there were some fixed buffers inside openssh that prevented copy operations from fully utilizing high speed network links…
$ read more →Gathering statistics from your postfix mail logs
I have been supporting postfix mail relays for several years. When you are tasked with ensuring that your SMTP relays work without issue, you need to stay on top of the logs and the current operational state of each server to prevent major problems from occurring. Problems can take the form of blacklisted relays, mail routing loops, spam attacks or people trying to exploit a known issue in the SMTP software. All of these items suck for a site that relies on mail to function…
$ read more →Using netstat and dropwatch to observe packet loss on Linux servers
Anyone that is running a modern Operating System is most likely utilizing TCP/IP to send and receive data. Modern TCP/IP stacks are somewhat complex and have a slew of tunables to control their behavior. The choice of when and when not to tune is not always super clear cut, since documentation and the advice of various network "experts" doesn't always jive. When I'm looking into performance problems that are network related one of the first things I review is the netstat "-s" output: The netstat output contains a slew of data you can be used to see how much data your host is processing, if it's accepting and processing data efficiently and if the buffers that link the various layers (Ethernet -> IP -> TCP -> APP) are working optimally…
$ read more →