Monitoring and analyzing performance is an important task for any sysadmin. Disk I/O bottlenecks can bring applications to a crawl. What are IOPS? Should I use SATA, SAS, or FC? How many spindles do I need? What RAID level should I use? Is my system read or write heavy? These are common questions for anyone embarking on an disk I/O analysis quest. Obligatory disclaimer: I do not consider myself an expert in storage or anything for that mater. This is just how I have done I/O analysis in the past. I welcome additions and corrections. I believe it’s also important to note that this analysis is geared toward random operations than sequential read/write workloads.

I thought I would write a little about Office 2010. I realize this isn't about Windows, but Office is important too. I really think that Microsoft messed up the user interface on 2007, and I know a lot of people that feel the same way. Granted, they were looking to improve the user experience, and they primarily had the new user in mind. But what about the literally millions of users that they left behind?

Well, it looks like the interface has been improved further for 2010. The ribbons are still there, but at least it appears to look a little bit more like the 2003 interface than 2007 did. One aspect of 2010 that really gets me excited is the Office Web piece! This looks like it might be the start of a beautiful friendship between a lot of new users and Microsoft, especially because the web access is free to everyone. It includes the abiliy to store documents online so the user can access them from any computer that has access to the Internet.

In my last post I gave some first impressions using bcfg2. We ended up selecting bcfg2 as our configuration management system for a new project. I'm going to start with the "bad" and get that out of the way.

The bad

The biggest problem so far is getting the entire team up to speed on bcfg2. It's very different from what we're used to with cfengine2. One challenge is wrapping your mind around what I think is one of the basic tenets of bcfg2 - you specify everything twice. Let me explain: bcfg2 separates out configurations into an abstract specification, and a concrete implementation. For example, you may decide a certain group needs an /etc/motdfile. So in the abstract configuration (called a Bundle), you simply say: <Path name='/etc/motd'/>. This is great for easily seeing what should be there. However, it gives no clue as to where that file comes from. It could come from the plugin that distributes static files, or one of the plugins that parses templates and distributes the results.

I picked up a couple of the new WD Green 1.5TB drives recently. I needed some new storage, and these drives have a lot going for them:

- 64MB cache - low power usage - 4K sector size

I knew going into this that the 4K sector size was proving problematic in some circumstances. In short, these drives have a 4K physical sector size, but present 512B logical sectors for better BIOS/OS compatibility. All well and good, except this can cause partitions to be misaligned and lead to writes falling across different physical sectors. This article discusses some of the performance implications. I verified on my drives that properly aligning the first partitions does lead to >3x performance increases.

At $WORK, we currently use cfengine 2 to manage on the order of 300 systems. It works, it does most things we want. We use the Singlecopy Nirvana "pattern" to distribute configuration files of all types, have copious shellcommands, and even a few editfiles. In general, it does what I need, but not necessarily the way I want. The general structure of our cfengine configurations is (for the most part) unchanged since I implemented it ~4 years ago. We've done the necessary things to support new OS's, new architectures, what have you, but that's about it.

Like many of you, I like to have some of the latest technology at my fingertips. Specifically, I'm talking about my workstation. No, while I don't have über-fast sexy hardware, I like to at least have the latest OS flavor(s) installed so I can play with new features.

The other day I decided it was "time" for Fedora 12. Time to play with ext4.

Before I begin, a small word about my prior setup:

disk 1: Windows XP, NTFS partition. Old, crusty, not used anymore, as I have XP in a VM now.
disk 2: Fedora 11, Linux LVM partition. No real data of value, but it does house my XP vm.

Coming from a rather hardcore X11 background (mm, X10 with b/w displays), my adjustment to Life With Mac has been made much easier recently with the installation of a utility called Zooom/2. [2] Although it can't (yet) bind my X11 finger-macro of alt+mouse2+move to window-move, it can bind it to any combination of modifier keys, defaulting to command+shift+move. window-resize is bound to option+shift+move.

Along with autoraise and focus-follows-mouse capabilities, this utility has already earned a permanent spot in my essential Mac folder.

A small dampener: There's still a few bugs or design interactions lurking about, especially when combining autoraise, focus-follows-mouse and Spaces. Focus can shift bizarrely to another space when least expected, and some applications get upset about abrupt resizes and can abort. The developer has been quick to respond to my bugs reports thou, so I currently have hopes that these issues can be addressed.

It's Friday afternoon, about 2pm. I have a console/management server that is having some difficulties with Splunk 4. Whether or not the RHEL version (4.8) is the cause-- I have finally decided it's time to upgrade to RHEL current (5.4). That way I can at least eliminate the OS as the cause of the issues I'm seeing, and so on. And, being Friday afternoon, it's quiet.

What could possibly go wrong?

I checked my ISO directories for the particular RHEL image I need (ver 5.4, x86). Yep! I've used it a lot on several VM's, but haven't needed to upgrade anything that's actual hardware yet. No problem, a few minutes later I had a hot & crispy DVD in my hand with 5.4 freshly burned on it. Off to the data center.

As our technology grows, along with pieces-parts from multiple vendors, so does our job as the technical equivalent of Dr. House. In many ways, what we do as sysadmins is very very similar to that popular TV show (which, by the way, is one of my favorites - his crankiness is something I intend to aspire to...). On the TV show, House guesstimates the problem with the patient of the week-- administers various types of remedies (all of which fail), the patient nearly dies, and at the last second- in a brilliant flash of insight, he figures it out.

Today is a day like many other-- I'm going through the above process. Software: Sun Storedge Enterprise Backup. Backup server: Sun X4500 ("thumper"). Tape Drive: Sun Storedge SL24 tape loader.

We are migrating from our old mail server to a new one with a completely different internal structure and many new features. I am using imapsync to migrate the data over (slow, but it works well). Before we do the last sync, my boss asked me to figure out a way to run the servers side by side so we could have the early adopters try the new mail server. The kicker is that both servers are set up to serve the same domain and we need to be able to have the early adopters get to their mail from either server.

This solution works for us, but I am not sure if this is the best solution and would love to hear your ideas.