June 3rd, 2010, 12:43 hours
After two more happy years on a dedicated box at Liquid Web, I’ve just finished up a test run on a new “cloud” environment and I must say… I’m sold!
The new Storm service is Liquid Web’s answer to the demand for a stable, reliable cloud, without the caching and performance issues that have plagued many other’s attempts (*cough* Mosso *cough*). Rather than setting up several web servers, several database servers, and several file servers, then patching them all together and hoping the result is useful, Storm’s approach is to create the cloud between the hardware and OS layers. I quote from the Storm site…
Storm On Demand’s stated server resources are entirely allocated to your server environment and will not be affected by the performance of other servers. Storm On Demand CPU’s represent a full physical processor core and are not shared with any other users. The distributed nature of our cloud architecture allows you to rely on the performance of your servers to be consistent and in line with what you would see with a physical dedicated server.
Unlike other cloud hosting providers, Storm On Demand does NOT utilize bursting shared resources or centralized storage which can have a significant negative performance impact for others utilizing the same resources. This helps ensure that Storm On Demand customers will always be guaranteed the performance that they requested from their server and will always get their full value.
The result is cacheless goodness, root-level access to dedicated resources, and the familiar cPanel/WHM management interface (optional). In practice, the end product lives up to everything the “cloud theory” promised us in the past and offers vast improvements in responsiveness, reliability and cost-effectiveness over traditional dedicated servers.
For a full list of features and specifications, or to make the switch, go check out the Storm site.
Tags: cloud, dedicated, host, hosting, Liquid Web, server, servers, site, Storm, Storm On Demand, web, website
Posted in Business, Design & Web, Geek & Technology | No Comments
Internet Explorer 7 was great. A lot of folks will hate all things Microsoft to their grave, but I have to give them credit where credit is due. IE7 worked, and it worked good. It was finally easy to make “The Big 3″ all render a page in almost exactly the same way. I thought the long struggle was finally over. Then Firefox 3 came out and completely ruined the concept of an auto-cropping <div>.
Microsoft has decided to screw up a good thing as well in an effort to out-muck the folks at Mozilla, or as they would put it “become more standards-compliant”. They have once again changed the way margins, padding and the box-model in general work in IE8 (causing many an animated menuing/accordion script to fail). Fortunately, IE7 isn’t gone, just muzzled. Adding the following meta tag (at the top of the meta section, or IE8 may ignore it) will force IE8 to render the page with IE7:
<meta http-equiv="X-UA-Compatible" content="IE=7" />
It is little more than a band-aid approach, but it works, and in my case buys the time I need to find a real fix for several production sites this issue affected. My thanks to Estelle for her post on this solution.
No Comments »May 7th, 2010, 11:53 hours
January 12th, 2010, 14:18 hours
ChazzLayne.com has been plagued by an intermittent error that on IE7 causes the page to vanish, and on IE8 causes it to stop loading wherever the error occurs (usually resulting in half the content and the entire sidebar disappearing). “Details” on this error, vague as usual, can be found here: KB927917.
The Snowball
I had time to do a little digging today, and through the WordPress.org plug-in repository was able to track the problem down to the beloved Flickr Tag (a better WP-to-Flickr plug-in there exists not!), but fortunately the trail didn’t end there. A comment mentioned there being a problem with Lightbox, and that disabling Flickr Tag’s Lightbox setting would fix the problem. Sure enough that worked, so naturally I set about to fixing Lightbox.
I started at the beginning – thinking that the copy of Lightbox included with the aging-but-trustworthy Flickr Tag must simply be out of date I set the plug-in to simply output Lightbox-ready HTML, then installed the latest version of Lightbox on my site. Feeling victorious I opened up IE and pointed the browser to a Flickr-heavy page on my site only to watch the whole thing simply go *poof*, just as it always had. Of course, the support forums at Lightbox proved to be little help. The only reference to the problem was a thread where someone got close to the answer, but mistakenly blamed the problem on a conflict with a Twitter plug-in and considered the case closed. It was then I realized the out-of-date Lightbox package included with Flickr Tag also included yet another out-of-date package: Scriptaculous.
The Cure
Sure enough, included in the latest newest bestest downloadable package of Lightbox 2.04 is an out-of-date copy of Scriptaculous that triggers the KB927917 bug whenever you look at it wrong. With Scriptaculous 1.8.3 installed, I can now enjoy my favorite, unsupported Flickr Tag plug-in for WordPress – a plug-in which still works flawless despite no updates having been released since WP 2.7. If only everyone wrote code this well…
Tags: 7, 8, 927917, bug, flickr, Flickr Tag, IE7, IE8, Internet Explorer, KB927917, Lightbox, plug-in, Scriptaculous, wordpress
Posted in Design & Web | No Comments
I just found a neat little Scriptaculous-based slideshow when I was looking for something to replace the Flash-based one in my portfolio. This is the first time in a long time that I’ve gone to install someone’s application and not had to do anything but copy+paste to make it work. In theory, I’m guessing it would probably handle more than just images…
No Comments »May 24th, 2009, 17:47 hours
April 20th, 2009, 18:34 hours
From time to time, I get curious who’s using my pictures and how they found them. Today I was looking through my stats on Flickr and was surprised to find some of my pictures in the top 5 search results for the following keywords…
#1 spot: “hoover dam pill box” on Yahoo! Images
#2 and #3 spot: “Busse Game Warden” on Google
#2 spot: “laguna indian” on Google
#3 spot: “kel tec p32” on Yahoo! Images
#5 and #6 spot: “carbonate canyon” on Google
January 24th, 2009, 19:03 hours
No, not the Hope & Change kind of change. I’ve re-organized a few things around here and added a couple sections. Since BAT*21 is now up and running, I’ve copied all the Gear & Skills posts (and all the blade-related Guns & Knives posts) over to that new site. The old posts will still be available here, but I will be posting any future gear reviews I do as a part of that project. Since that leaves nothing here from those topics but firearms, I’ve created a new section to replace Guns & Knives called… Firearms.
Since moving to the middle of nowhere, I’ve also been doing quite a bit of maintenance and modification to my Discovery on my own so I added a section called Land Rover for any useful tidbits I find working on that project, and any post-worthy off road trips.
fotikus is also up and running now, though still under construction. I’ve only had time to put a hundred or so photos up there, but I expect to have the rest done before the end of the month. Pictures on fotikus are free for non-commercial use, or $25 each for commercial customers.
Tags: BAT*21, change, ChazzLayne.com, Discovery, Firearms, fotikus, gear, guns, knives, Overland, photo, photography, photos, picture, pictures, posts, reviews, Rover, sections, site, skills, survival, topics
Posted in Business, Design & Web, Firearms, Personal | No Comments
April 23rd, 2008, 11:09 hours
Some of you may have run into a rather frustrating issue when someone on a Mac emails you a JPEG. It will not open, at least not in any of the built-in viewers (Photoshop doesn’t much care). The problem lies with a coding error in Entourage (the Mac version of Outlook) and the way it handles file attachments that wasn’t really noticeable until a recent security update for Outlook. As it turns out, this appears to be the same problem some people have also been having on *nix-based systems. I’m sure the good folks at Microsoft are working on a patch for the error on the Entourage side of things, but it could be a wait as it seems they themselves didn’t even know about it until last week.
In the mean time there are three options available for a work-around, depending on your situation. You can remove the security patch from your Windows-based systems and leave yourself open to attack, use a different email client on your Mac, or transfer JPEGs from your Mac another way (which works fine as this is only a problem with Entourage). Removal of the security update is quite easy, just open your Programs and Features control panel (Add/Remove Programs on XP) and look for the update near the top titled “Security Update for Microsoft Office Outlook 2003 (KB945432)” published on March 11th. You might need to have your Office CD handy to remove this update.
Of course, if most of your clients are on Windows the first option won’t do you much good.
It would seem there’s a setting different here than the old host, which Binary Blue does not like.
Some of the other plugins are spewing gibberish too, back to the drawing board…
1 Comment »March 21st, 2008, 21:12 hours
March 21st, 2008, 20:37 hours
Well, that was a complete failure. As happened last time I followed the temptation to leave and cut costs, I find myself back at Liquid Web (again) with an even more expensive “hosting plan” (again). As the saying goes, “you get what you pay for.” So, I’ve moved off the 2.8GHz P4 with a gig-o-ram and single 80-gig drive that was fast approaching 3 years of age (the age at which you start to fear drive failure). In it’s place sits a dual-core 2.6GHz box with a whopping 4 gigs-o-ram and twin 160-gig drives in a hardware RAID1 (a “mirror” in English, so in 3 years I wont fear the failure again). Accommodating as ever, Liquid Web managed to swing the replacement in for no more than the cost of the RAID, and even threw in a third 160GB drive. They’ve also made some significant improvements in cPanel/WHM and available security options in the three years since I’ve looked at managed servers.
Mosso, which sells itself with a convincing pay-as-you-consume auto-bill-your-clients Windows-and-Linux-at-the-same-time cluster > server argument, looks quite professional on the outside and at a mere $100+growth monthly sounds quite the good deal. Inside it’s a completely different story, however. Everything about the place seems very jerry-rigged and downright amateurish. In fact, having given it a go for almost two months I find myself questioning their claim that they really have the official backing of RackSpace (as their about-us story says) or are even a RackSpace venture (as their footers say). I find it far more likely that it’s a couple of RackSpace employees (as they’ve claimed) that just wrangled a bunch of free or seriously discounted servers from the company so they can play host. There were little troubles here and there, and constant chats to support to have them do something their shoddy management interface errored-out doing (which grew tiresome), but it really came to a head when suddenly all of my sites, my clients’ sites, and for that matter all php sites went down for a day. I mean really, how in the hell does a serious host let a single failed power strip bring down the entire PHP section of the cluster for more than a day? THERE’S A REASON COMMERCIAL GRADE SERVERS HAVE AT LEAST TWO POWER SUPPLIES AND TWO POWER CABLES YOU IDIOTS!!! As if that wasn’t enough, the way they have their setup structured left all sites intermittent at best for the next several weeks. I could understand if this was a one time thing, but inquires answered by long-time Mosso customers assured me that I could expect much the same to happen every couple months. No, I very much doubt RackSpace has any more than a passive interest in Mosso, and even then only so much as they don’t set the building on fire.
The other good thing to come out of this, I suppose, is that the move from Liquid Web (cPanel) to Mosso (proprietary) forced me to get off my duff and rebuild the template for Chazz.US on the latest WordPress – something I’d been putting off since before Widgets. Of course, I no later get this (almost) done and they release RC1 of the heavily re-developed WordPress 2.5.
Tags: amateurs, cluster, dedicated server, design, failure, host, idiots, Liquid Web, Mosso, RackSpace, redesign, server, wordpress
Posted in Design & Web | 1 Comment
February 16th, 2008, 12:04 hours
Moving to a new host is just so much fun, especially when it’s a dedicated server with a few dozen “accounts” and the new host runs on different technologies. Still, the worst appears to be over and I think we’ll be finished up this weekend.
As I mentioned a while back, the easiest way to accomplish this is with shell access to one of the servers and NcFTP. I don’t have that luxury at the new host, but fortunately this method works just as well with the put command as the get command.
put -R *.*