Forcing a Remote Citrix Server to Use a Local Host Cache File

Posted by Carson       Trackback

At a previous employer, we used Citrix Metaframe Servers at four different sites, with one of the sites handling the central configuration of the Citrix Server Farm.  This worked great for the most part, but during extended MPLS WAN (Wide Area Network) outages, the remote servers started to hang up, as they could not “phone home” to get the central configuration datastore.

Forcing the remote servers to temporarily look only at their local configuration cache would result in much better performance.  Once the MPLS WAN was back up, I would need to revert to the standard settings.

Part I: Use local cache on when the central datastore is unavailable
1. Stop IMA Service on affected remote server
2. Update registry entries:
HKLM\Software\Citrix\IMA\DatabaseDriver = IMAAccess.dll
HKLM\Software\Citrix\IMA\DataSourceName = M:\Program Files\Citrix\Independent Management Architecture\Imalhc.dsn

3. Start IMA service

Part II: Use central datastore for configuration (once WAN is back online):
1. Stop IMA Service on remote server
2. Update registry entries:
HKLM\Software\Citrix\IMA\DatabaseDriver = IMASql.dll
HKLM\Software\Citrix\IMA\DataSourceName = M:\Program Files\Citrix\Independent Management Architecture\MF20sql2k.dsn

3. Start IMA service


Tech Notes!

Posted by Carson       Trackback

I was cleaning up Outlook recently in an effort to be more organized when I realized that I have accumulated a lot of personal technotes over the years, and I decided it would be a good idea to publish some of the more useful topics.

I’ll start by posting a new article tomorrow (time willing), and I’ll continue to try and post a couple of notes every week until I’m all out.  Topics will range from Windows XP to Cisco to Windows Server to RedHat / Debian Linux.  t’s a pretty wide range of topics, but it’s all stuff that I felt was worth noting at one time or another over the past several years.  Hopefully someone else can get some use out of this stuff as well.


The ISP Upstream Dilemma

Posted by Carson       Trackback

Comcast just announced that they have increased their premium tier high-speed internet package from 8mbps to 16mbps. It seems as though we might be finally catching up with the rest of the world in terms of real “high-speed” internet access until you read the fine print and discover that Comcast has not increased the upstream speed – it remains at 768kbps for even the premium package users, and 384kbps for standard users.

It’s puzzling that ISPs all seem to limit the amount of upstream data so heavily. About 10 years ago, as DSL and cable modem service began to proliferate, the standard data package was about 1.5mbps downstream and 768k upstream. The general idea was that the ISPs did not want individuals hosting web servers and other applications on their networks, so they limited the upstream pretty heavily. That seems reasonable, until you jump to present day, where the downstream is on average 3-5 times faster, but the upstream is the same or SLOWER than it used to be a decade ago!

With the advent of digital video and 10+ megapixel cameras becoming the norm, along with amateur video casting and web conferencing fast becoming mainstream, the slim 384kbps upstream pipeline just can’t keep up. Those that are a little more tech savvy and understand this limitation can take steps to live within the tight space of upstream bandwidth – things like shrinking large pictures and dropping the webcam quality can help conserve bandwidth. However, grandma just can’t understand why it takes nearly an hour (if successful at all) to send an email to her friend that has a half dozen 10 megapixel pictures attached.

The old “hosting web servers” argument just doesn’t hold up anymore. ISPs should stop gloating about their so-called high-speed internet access and increase upstream data to at least 1.5mbps to accommodate the online lifestyles of today’s internet user. Surely the backbones that the ISPs use can handle the additional speed. All internet backbone lines, whether they be older DS-based or current OC-based lines, utilize symmetric link speeds, both upstream and downstream, unlike the heavily lopsided end-user internet connections.

UPDATE: Comast recently announced faster uploads for all customers at no charge…they must be reading my blog!  Standard tier Comcast customers now get 6mbps down / 1mbps up and the premium tier customers get 8mbps down / 2mbps up!  Finally!  A nice speed bump…posting pictures to the family album is loads faster…


Where Are All the Moon Colonies?

Posted by Carson       Trackback

I was watching some show on the history channel about integrated circuits, and they mentioned that the Furby had more processing power than the computers on board the NASA Apollo 12 spacecraft. For some reason, this statement grabbed me and sent my mind off on a rather wild tangent.

I started thinking about the Apollo 11 mission, in which we (Americans, that is) put a man on the Moon in 1969. Nearly 40 years ago, Neil Armstrong is hopping around on the lunar surface, probably having the absolute time of his life. I can only imagine the future that space-loving hopefuls imagined after this momentous event. There were a few more missions to explore the surface of the Moon over the next few years, until the Moon missions rather abruptly ended in 1972 with the Apollo 17 mission to the Moon.

Thirty-six years later and there have been literally no advances towards life on the Moon. We’ve shrunk the computing power of the Apollo 12 spacecraft down to the size of a dime and made it cost-effective enough to place in a ridiculous toy, and yet we’re still stuck on terra firma. Certainly if we can accomplish such a feat, we could miniaturize life support systems to a cost and size reasonable enough for use on the Moon.

I suppose in the case of life on the Moon, much of it boils down to politics and money. Apparently, the Moon doesn’t have enough natural resources to support colonization. Too bad there aren’t large oil reserves or something along those lines that Exxon could exploit….with 40.6 billion dollars in profits for 2007 they certainly have enough money to at least consider the feasibility of exploration missions and mining operations…


Fun With Windows SMB Share Migrations

Posted by Carson       Trackback

Let’s say that you get funds approved to put some new Windows file/print servers in place. While planning the migration, you have a few simple requirements:

1. You want to migrate files and printer config to a new server
2. The new server will have a new hostname and IP address (in my case, I wanted to get rid of underscores in the hostname)
3. The old SMB paths and shortcuts must continue to work so users don’t panic

Seems easy enough – a simple DNS alias ought to do the trick, right? Not quite. While a DNS alias will allow clients to resolve the old hostname to the new IP address, and simple operations like icmp requests will work, access to SMB shares on the server will return “A duplicate name exists on the network” errors.

According to Microsoft, this occurs because Windows is not “listening” on the new alias! I’ve never seen anything like this with any Linux or Unix servers I have aliased, so I was a bit perplexed by this behavior. Fortunately, the problem can be eliminated by adding a registry setting which will allow SMB shares to work via the alias. However, the setting does require a reboot, which is a bit of an inconvenience.

1. Start Registry Editor (Regedt32.exe).
2. Locate and click the following key in the registry:


3. On the Edit menu, click Add Value, and then add the following registry value:

    Value name: DisableStrictNameChecking
    Data type: REG_DWORD
    Value: 1

4. Quit Registry Editor.
5. Restart your computer.

After thorough testing, I rolled out the first of many new servers, and to my delight, SMB access to the new servers works flawlessly! Users can still have shortcuts to the old server and SMB paths and everything is redirected properly. As planned, the changeover to the new server is completely transparent to end users.

MS KB281308


Skullcandy: Cheap Earbuds That Don’t Suck

Posted by Carson       Trackback

Skullcandy Ink’d Smokin’ Buds are pretty decent, considering they’re only $15. My ipod headphones went bad recently so I picked up a set of these earbuds, mainly because I was amused by the name “Skullcandy” – plus they were amongst the cheapest ones I could find.

I thought that they would be pretty horrible, but they actually sound great! The bass response is excellent, although highs are a bit lacking. I don’t have much to benchmark them against, but my ears like ‘em and so does my wallet.

Skullcandy Ink’d Smokin’ Buds


DRAC5 is your pal.

Posted by Carson       Trackback

Remote server administration can be a real pain, especially if you don’t have a knowledgeable onsite person to help troubleshoot problems. In my current role, I am responsible for 11 remote sites which all have servers that occasionally need maintenance or troubleshooting. Fortunately, I just finished rolling out new servers to all of these locations and remote administration is a whole lot easier now.

All of the sites are now running Dell 19XX series servers, and every new server that went out included a DRAC5 card, which is specially designed to work with the PowerEdge 19XX servers. I figured that these cards would be a good idea for remote sites, but they work so well that I have since purchased a bunch of extra DRAC5 cards and installed them in all the servers at my local site as well, which saves me trips to the office on occasion.

The Dell Remote Access Controller version 5 (DRAC5) cards are basically a tiny system-on-a-chip in the form of an addon board that connects directly to the server motherboard and power supply, and has an external RJ45 network jack. They are completely independent of the main server chassis and even if the server is off (but plugged in), the DRAC5 card is on and will respond to network requests. This makes it possible to do stuff that would normally only be available locally. Someone still needs to plug the server in and connect it to a live network port, but that’s pretty much it.

The DRAC5 card does a lot of cool stuff, but I have found these features most useful:

1. Remote power off/on: As long as the server is plugged into a power source and the DRAC5 card is plugged into a live network jack, the server can be turned ON / OFF remotely. It can even be forced to turn off in the even that the server freezes. It is essentially the same as hitting the power button on the server.

2. IP KVM: Dell calls it a remote console, but it is a full functional IP KVM connection that runs thru the DRAC5 network interface. This means that even if the server LAN ports are not plugged in, you can still get a remote session where you can interface with the Linux console or Windows desktop.

3. Virtual Media:
In addition to the CD/DVD drive in the server, a virtual device is created and you can put a CD in your workstation and mount it on the remote server. Even more nifty is the ability to directly mount an ISO image!

Used in combination, these three features can be extremely handy. It is possible to ship a bare server with no operating system to a remote site and perform the entire OS install remotely. First you would connect to the DRAC5 web management interface and launch the IP KVM, then flip over to the virtual media options and mount your Linux or Windows ISO/CD/DVD. Finally, you would power on the machine and watch it post in the IP KVM, then start the installation!

Suffice to say, I doubt I will ever roll out another Dell server without a DRAC card installed.


Pet Peeve: Streaming Video Players With No Volume Control

Posted by Carson       Trackback

What kind of idiot web admin serves up streaming video without volume level control? It really pisses me off when my only sound options are ON / OFF, especially when the fixed audio volume is way too loud.

My rating for the consumer reports website is: CRAP!



Zelda Makes You SHOUT!

Posted by Carson       Trackback

The Nintendo Wii gets a lot of attention due largely in part to it’s innovative wiimote, and rightfully so – it’s a completely different experience than playing other consoles. It seems everyone from little kids to grannies love to play the Wii. However, I still log far more hours on my DS lite (probably because of the portability factor). Like the wiimote, the DS Stylus allows for some revolutionary yet simple game play.

My latest obsession has been The Legend Of Zelda: Phantom Hourglass. I was a bit concerned at first, as you have to make Link move around with the stylus. At first I was annoyed by this, and some times it still bugs me when my hand gets in the way of the screen, but overall the controls are decent. What has really drawn me into the game is the really unique puzzles the methods you have to employ to solve them. Nintendo has constructed some really innovative ways to use the DS lite for game play. I’m really not all that far into the game and have already come across three pretty tricky puzzles (caution – these are sort of puzzle spoilers):

1. Blowing out the torches at a temple entrance- First I tried swiping the candles with the stylus, then I realized that you LITERALLY HAVE TO BLOW on the DS lite to blow out the candles! The built in microphone senses you blowing on the it and the candles go out.

2. Yelling at the guy in the machine shop to give you a salvage arm for your ship- I had to YELL OUT LOUD to the character, “I WANT THE SALVAGE ARM, PLEEEEAAASE LET ME BUY IT!!”. I was a bit timid at first and I ended up having to yell three times because the dude said I wasn’t enthusiastic enough the first two times. Even after three attempts, I must not have yelled loud enough, cause it still cost me 1000 rupees. As fate would have it, this particular part of the game transpired while I was parked on the toilet.

3. Pressing a spot on a map to the your sea chart- This one had me baffled for the longest time. I left the temple a few times and was steaming mad because I couldn’t figure it out. The game kept hinting that I needed to press the map to the sea chart. I thought I was going to snap the stylus in half as I kept pressing on the map. The game went as far as to tell me that the maps were like a mirror image. I was starting to get pretty pissed off at this point. Finally I realized that since the images on the top and bottom screens were mirror images, if there were some way to drag the top screen onto the bottom, I could press the maps together….it was then that my brain clicked out of 2-D video game mode and I realized that if I actually closed the DS lite lid, it would be like pressing the maps together. Normally when you close the screen of the DS lite, it goes into a suspend mode, but in this case when I closed and re-opened it, the game registered the action and the marker was pressed onto my sea chart! What a cool way to take advantage of the hardware!

This is the sort of stuff that makes Nintendo and their game developers stand out above all the other software / hardware players. I’m constantly amazed at the cool stuff the folks at Nintendo cook up. Things like the examples above are what keep me hooked on my DS lite, wondering what crazy stuff I’m going to have to do in order to solve the next big Zelda puzzle!


Akismet Stomps Spam

Posted by Carson       Trackback

I had no idea how much comment spam Akismet had caught on this site until I logged in to check a couple of obvious spammer trackbacks:

Akismet stomps spam

13,223 blocked spam items…that’s quite a bit for a blog that nobody reads!

« Previous Page  Next Page »