A post today on Slashdot got me thinking about advertising. Specifically advertising on the internet, but also advertising in general.

The article linked off the Slashdot post was written by a man named Jim Lynch, a long time writer in technology media both digital and print. Mr. Lynch is apparently annoyed by a new feature in Apple’s just-released Safari 5 web browser called Reader. Reader is a feature that, when selected by the user, attempts to detect “article” content on a web page and display it in a simple format which is larger and often easier to read than the normal web site layout. It also attempts to detect multi-page articles and automatically display further pages as you scroll down, effectively creating a “print” view for sites which may lack such things.

What bothers Mr. Lynch basically comes down to advertising. When using Reader, if it works properly all ads are stripped out of the content. More importantly for some, the automatic loading of the next page means cost-per-impression ads get many less views as they would only show on the first page before the user clicked the Reader button.

I understand the key point behind his complaint, web sites cost money to run and that has to come from somewhere. This site costs me about $275 a year between domain registration and server space, and it’s fairly low volume (understatement of the century, I average less than 40 pageviews a day not counting spiders). I pay this out of pocket, since for my use the domain is for my email and the VPS is just a place for me to experiment. As far as I’m concerned I’d be paying for them both anyways, so why not put something there? Obviously that reasoning doesn’t tend to apply outside the range of personal blogs and the costs are much higher when you start talking real traffic levels requiring real servers rather than a virtual slice of one.

Unfortunately, I can’t help but not feel the slightest bit of sorrow for advertisers and those running advertising when they complain about their ads being blocked. They’ve for the most part brought this on themselves, by designing their ads to be as intrusive and annoying as possible. Web publishers have been just as badly a part of the problem, injecting ads as if they were content, allowing nuisance ads with autoplay audio/video or various popup/under/over windows, and in some particularly annoying cases using the content as the ad with IntelliTXT and the like.

We’ve already seen what the ability to skip ads has done to the television industry. For years they thrived on annoyingly loud and repetitive ads which seemed to rely on the “any publicity is good publicity” theory. As soon as the DVR became common the ad market pretty much fell apart on anything people weren’t watching live. Now that extensions like Adblock for Firefox and Apple’s new Reader are making it easy for the average user to dodge ads (rather than us geeks who have been doing it for years) the internet ad community fears the same thing happening.

All I have to say is that the internet ad industry needs to learn from the successful television ad campaigns.

First and foremost, DO NOT PISS OFF YOUR POTENTIAL CUSTOMER!!!!!!!!!
If an overly loud and annoying ad comes on the radio or TV, I’ll turn the volume down or change the channel if I don’t really care for what’s on while making a mental note to avoid the advertiser if possible. The same applies to internet ads. If your ad stretches over the content I’m trying to read, starts playing audio out of nowhere, makes half the words on the page pop up product links, or otherwise interferes with my reading of the content I will go out of my way to avoid your product where possible. If ad blocking is available, I’ll turn it on immediately when any of those happen and may make a note to avoid the site where it was seen as well.

Second, draw my eye the right way. You do not have to be loud, either literally with audio or figuratively with bright/flashing colors. Use your space to make me interested in what you have, then if I actively click on it you can load your content of choice. This is more for advertiser rather than publishers, but due to point one publishers would do well to enforce point two.

Third, be relevant. If I’m reading a site about cars, an ad for purse built to carry small dogs is most likely irrelevant. Again this is for both publishers and advertisers. Ad networks which do not target based on content are outdated and should be dropped immediately from both sides.

Fourth, don’t try to shove too many ads in my face. I myself start getting annoyed when there’s more than 3–5 ads on the screen at one time, depending on the amount of content and such. Sites that split articles in to a huge number of short pages in order to increase impressions for ad purposes fall in to the same category (and I believe these sites are the greatest reason for the Reader feature). Dividing articles in to multiple pages is fine, but don’t do it unless you have at least as much information on a page as an average magazine. Two paragraphs and a few pictures are not a page.

The short version is provide ads that don’t annoy the reader and preferably are something they might actually want and you won’t have as many blocking them. If the relevance goes up, more people will click on them too. As for the rest, those who have already decided to install full ad blockers, those are gone already. You won’t get them back, it’s just too nice. Download Firefox, install Adblock Plus, and subscribe to one of the popular filter lists like Easylist. Now turn it off and browse to a few popular news sites. Turn it back on and reload those pages. If you don’t agree that this is a much cleaner and more enjoyable way to browse the internet you’re blind.

If you are seeing this post, your DNS servers have updated and noticed that my old VPS on JaguarPC is no longer where they should look. My blog and other random shit has now been moved over to Linode, where they don’t block IRC and other things I run on my box.

This morning I rebooted my test box running VMware ESXi 3.5 to complete the upgrade from Update 3 to Update 4. The hypervisor came back up, but no guests were running and when I popped open the VI Client it indicated that there were no datastores configured and it could not find any of the virtual machines I had in inventory. It saw the internal disks and that they were formatted VMFS, but would not allow me to do anything other than format them over again.

Normally this would have simply annoyed me since I would have lost my test VMs, but they don’t take long to build so I’d have just formatted them and gone on with my day. Unfortunately within the last week we had temporarily moved a critical application’s VM to this box and we had not properly reconfigured backup. I could restore from the week old backup, but there would be hell to pay.

Since the VMFS partitions were clearly visible I felt I had a chance, but I’m still new to ESX/ESXi so my first step was to flip over to my always running irssi session (if you use IRC and do not use screened irssi, go Google it now and enjoy) and ask for help in #shsc and #vmware. #shsc always has a few guys who work on large VMware installs idling, and of course #vmware is obvious. While waiting for any input from IRC, I went to Google for my next step. I knew ESXi has the capability to be accessed via SSH, but it’s disabled by default, so I looked up how to turn it on. A few minutes later after bringing a monitor over to the machine and rebooting it I had SSH access and could go through system logs from the comfort of my laptop.

In /var/log/messages I found two entries referencing my SATA controller which looked interesting:
May 5 14:34:35 vmkernel: 0:00:06:39.406 cpu0:3616)ALERT: LVM: 4482: vmhba000:0:0:3 may be snapshot: disabling access. See resignaturing section in SAN config guide.
May 5 14:34:35 vmkernel: 0:00:06:39.408 cpu0:3616)ALERT: LVM: 4482: vmhba0:0:0:1 may be snapshot: disabling access. See resignaturing section in SAN config guide.

This information, after a quick trip to Google, led to VMware’s SAN configuration guide which references similar issues occurring on SANs, so I tried enabling the resignaturing option and magically my datastores reappeared. After renaming them back to their original names and turning the resignaturing option back off I had all my data and was able to download the disk images and VMX files so I was safe in the event of a major problem.

At this point, I could see my VMs but the VI inventory was still convinced that they were on the “old drives”, so after a bit more time on Google I discovered the Import feature within the datastore browser and I was able to bring the VMs back in and get them booting up.

Screenshot showing my datastores and two VMs running

After confirming that the VMs I really needed were booting and operational, I shut everything down to move the server back to its spot in my rack. Fortunately everything came right back up so the pressure was now off.

Now my concerns shifted. If this happened once, what’s to stop it from happening again? I needed to figure out why it happened. Fortunately at nearly the exact moment I started thinking about this IRC came through for me. “jidar” in #shsc linked to this thread on VMware’s forum with literally the exact same symptoms. A few posts down was a link to this page which again matched my experience exactly and says that U4 updated a number of SATA drivers including the one for the ICH9 controller in my PowerEdge and changed the way they appear to the hypervisor, which led to it not recognizing the drives for what they are.

Right now I’m moderately annoyed at an update that’s not even enough to earn it a minor version number bump on a piece of software intended for enterprise use having a change with the potential to cause this, but on the other hand I don’t expect anyone who really cares about reliability to be using SATA local storage. Ah well, I learned a bit about navigating around ESXi’s internals.

Turned off the Twitter digests, those were just silly and annoying.

Fixed the pretty URLs and RSS feed (blasted mod_rewrite), added Gravatar support, a tag cloud, and a download system that I may or may not ever use.

The Subversion install/upgrade method wins again, bumping up to 2.7 was a single command while logged in as www-data on my server and then accesing wp-admin/upgrade.php

In his recent Home Media Server Guide, Brian Won of Ars Technica briefly discusses the heat and space issues involved with packing lots of hard drives in to a personal server. Since these problems are handled nicely by cases such as the Cooler Master “Stacker” line and the Antec Nine Hundred, I started thinking about power consumption.

My question then became whether it would make sense from a financial perspective to pay extra for larger hard drives which you could then use less of, thus burning less power.

The drives I chose to compare are the 16MB cache models of Seagate’s Barracuda 7200.10 family. These are popular pieces of equipment with good reputations and a wide range of capacities from 250GB up to the 750GB behemoth and an anticipated 1TB model in the near future. Choosing this family also simplifies things dramatically because the manufacturer’s specification for power consumption is identical across the entire line, at 13 watts average. Based on my own experiments with the 250GB and 500GB models using my trusty Kill-A-Watt, these numbers are fairly accurate.

With the power consumption being the same across the entire line, it’s clear that using less hard drives means less power consumption. This then shifted my focus from whether there will be power savings to how long does it take for the power savings to overcome the initial investment.

After looking at the price per gigabyte, I eliminated the 250GB model from this comparison. This is because it costs more per gigabyte than the larger 320GB model, thus requiring both more drives and a greater initial investment to provide the same amount of space. There will never be a return on the investment when choosing the 250GB model, and thus I only recommend them if you must have storage space right now and can’t stretch for the extra $15 to upgrade to the 320GB model.

Now, we’re left with the 320GB, 400GB, 500GB, and 750GB models. Using the prices from Newegg on 2/1/2006, they are 3.37, 2.86, 2.5, and 2.21 GB per dollar respectively. I’m going to aim for a total capacity of 1.5TB when choosing the number of drives for each size. This is partially because it’s easy to hit close to with all four sizes and partially because it’s a reasonable size for a home server.

• 5x 320GB @ $95 ea. — $475 — 1.6TB
• 4x 400GB @ $140 ea. — $560 — 1.6TB
• 3x 500GB @ $200 ea. — $600 — 1.5TB
• 2x 750GB @ $340 ea. — $680 — 1.5TB

As you can see, I overshot a bit with the two smaller drives, but this is as close as they’d get without having to push these numbers way out of the home server range.

Here in Toledo, OH, I pay $0.137 per kilowatt-hour for electricity. Forget the math, but at 13 watts a piece this means that every hard drive I don’t have to power in a 24/7 server is a savings of $1.30 a month on my electric bill.

When we look at the up front prices, I’d be paying $85 extra up front to cut from 5 drives to 4 with no difference in capacity, $125 extra to cut out two drives (and losing 100GB in the process), and a whopping $205 to bring it down to only two running drives (also losing 100GB). That’s quite the investment for such minor anticipated power savings.

By now it’s probably clear that the break-even point based on power savings alone is a long way off. I went ahead and did the calculations, and here are my results: The shortest break-even time goes to the 500GB drives, which would have paid for themselves in a few days past four years. Second place goes to the 750GB drives, which break even about four and a half months after the 500GB models. The 400GB drives are the real losers here, taking nearly five and a half years to break even.  SEE BELOW

Given the up-front savings, not to mention the flexibility you gain by using more drives (RAID levels 5 and 6 for failure-resistance), it seems that the only logical choice is to keep buying whatever size offers the best GB/$ bang for the buck. The money saved up front can cover the extra electric costs for years to come, and given the current rate of technology development by the time the larger drives would reach their break-even point there will already be single drives on the market capable of storing this entire example array many times over.

Note: It has been brought to my attention that the 400GB drives are currently selling for $120 rather than $140. This cuts the total price for a 1.6TB array of them down to $480, only $5 more than the 5x320GB array. With the savings of $1.30 per month, this initial investment is now paid off in just shy of four months, not to mention the advantage that four-drive RAID controllers are far more common than six-drive models, especially on motherboards. With that in mind, it’s obvious that the 4x400 array makes more sense from both a financial and practical perspective.