Bluetooth

I've got a handful of Bluetooth enabled devices. Some work, some don't.

Audio

The first Bluetooth device I bought was headset for my phone. I thought that it would make driving while talking a bit safer. TheHS850 didn't have very impressive audio, but using the headset with my phone was simple, easy and obvious. If the headset is turned on, the phone attaches to it and switches to the headset. If the headset dies or is shut off, the audio switches back to the phone hand set automatically and reliably. The phone has a simple and obvious menu to switch to & from the headset. It works.

I've bought an S9 to use with my phone just as well, mostly to use while mowing lawn. The audio is nothing special, but it pairs and connects easily to my phone, and give me adequate stereo fidelity with no hassles or issues. Pairing, changing tracks, answering the phone while listening to podcasts is simple and obvious. The buttons are way too touchy, but other than that, it works.

I also bought a Motorola T305 also, hoping to use it to play podcasts while driving on long trips. The concept is great, but unfortunately is simply isn't loud enough to overcome the road noise on my vehicles. Close, but not quite. My cars apparently aren't particularly quite. A T505 that outputs audio via your FM radio might be ticket.

On the computer side of the equation, I've got a handful of devices that support BT to some degree The Mac is pretty close to having decent BT audio. It doesn't automatically use the headset or headphones, unless I use a 'Use Headphones' menu, but once that is done, it uses the headphones when they are on, and switches back to on board audio when they are not.

The XP computers I have don't even have a usable BT stack built in, so I'm stuck using 3rd party drivers, with varying results. I'd say that they are barely functional. Vista is a bit better. The built in stack is usable, but it isn't a seamless experience for BT audio.

File Transfer.

Apple still treats Bluetooth file transfers like a 1980's FTP session. Of course you don't have to use a command line, but the GUI that they present is only slightly better. Apple's designers are are still thinking that you connect to a BT device, browse through the device file system, and transfer files to & from the remote BT device as you would a with 20 year old FTP session. Drag and drop works, but drag and drop is really still treated as a file transfer, not as a seamless part of the native file system.

Vista tries to treat a Bluetooth file transfer as a network attached drive. It builds folders in explorer for the various devices and lets be drag, drop, rename, etc. It doesn't seem very robust though. I can easily break it by playing around with my BT adapter (unplugging it and re-plugging it). It still thinks I'm connected, even why my phone is off, and generates useless error messages when I try to access the Bluetooth file system. It doesn't handle error conditions very well.

The same operation on my N800 is much different. The N800 sees BT devices as a mountable file system. Any paired and in-range BT file devices show up simply as mounted subdirectory that the file manger sees just like any other part of the N800's file system. That is one of the best parts of using the N800. If I see and interesting podcast, I save it to the N800, then drag & drop it over to my mp3 phone using the N800 file manager. It's not a separate application just for moving files around with Bluetooth, it is a Bluetooth file system seamlessly integrated in to the device. No tethering, no wires, no iTunes.

DUN


Nokia really did a great job integrating dialup networking with the N800's Bluetooth stack. If I want to use my phone for internet access from my N800, I run a simple wizard once, then pick the phone from the wireless connection menu. As far as the N800 cares, GPRS, Edge, 3G HSDPA, and B/G WiFi are all the same. And with 3G, my internet connection speed is around 1mbps. Very well done.

I played around with using Bluetooth DUN from both Vista and XP. With the proper software, it is pretty straightforward and simple. Half-way through editing this post from m Vista machine, I launched Motorola Phone Tools, clicked on the Internet button, and switched my network connection from Wifi to ATT 3G and finished the post. Pretty neat and easy.

Using Bluetooth for Internet access from the Mac wasn't anything like seamless or painless. Considering that the phone and the Mac are already paired, I obviously should be able to select my phone from the same pulldown as I select Wifi networks, like the N800, but following that thread leads to a dead end. Network preferences shows my phone, not connected (actually it shows a whole bunch of phones, because apparently it never occurred to an Apple engineer that I might want to delete a phone from network preferences......) and leaves me with no way of connecting . In any case, there doesn't seem to be a logical menu or button that connects my phone to the Mac via BT, in spite of trying multiple dialogues, buttons, pulldowns & whatever.

Bluetooth Preferences shows the phone, not connected, and leaves me with no combination of splat-ctrl-fn-alt anything that will light up BT to the phone. Using the BT browse device menu will let me connect to the phone via OBEX, but that doesn't light up the DUN connection. Network preferences shows all kinds of borked dialup connections, and does strange things when I try to configure them with Cingular's config.

In any case, randomly clicking around eventually gets it to work, somehow, but some unknown combination of magic clicks & incantations.

The Mac to DUN connection via a Bluetooth phone is a pain, simply because the UI isn't simple or obvious.

Presence

Instant messaging isn't about messaging. It's about presence.

I'll admit that even though I've had IM accounts for a decade, I rarely if ever signed in. I found the pop-up like nature of the common clients to be far too disruptive and annoying, and I simply preferred e-mail or voice for routine communication. Lately though, I'm finding IM, under the right conditions, to be far more than just a quick a dirty communications tool. I'm figuring out presence. 

If I want to communicate with a coworker, what's the first thing I do? Check IM. If they are around, and if they are available, I'll know right away. If they don't show up on IM, they set the 'do-not-disturb' bit and I know they are busy. If they are 'on call' but no near their computer, they set their status to 'Call my cell phone @ nnn-mmm-zzzz'.

It's presence.

What's next? 

I don't see any reason why things other than persons can't have presence. We've got a thousand or so network & server devices that really matter to us. People we provide service to tend to notice when the devices aren't happy. And we've tended to notice that when people that depend on us aren't happy, neither are we. So we like our devices to be happy.

So here's what I want my devices to do. When they are happy, they set their status on IM to Green & 'Happy'. When they are sad, because they don't have enough CPU, memory or bandwidth, or whatever, they set their status to 'Sad', and perhaps even 'Sad, need more CPU.' Now if I'm a person that either manages or depends on these devices, I build myself a buddy list populated with the devices (servers, routers, firewalls) that I care about, and I see right away if they are green & happy, yellow & sad, or red & dead. 

My Mac can even make funky noises when my devices transition from happy to sad  & back again.  So when my devices change status, they can tell me. Nicely 

Red power, Blue power

While building out a new data center, we took a look at what our practice had been for provisioning power for servers and devices within a rack. As pretty much everyone does, we bring two separate 220v circuits into each rack. One circuit terminates on some kind of power distribution unit (PDU) on one side of the back of the rack, the other circuit terminates on a different PDU on the other side of the rack.

When we provision a new rack and install equipment into the rack, obviously we carefully route and label all the power cables and make sure that each power supply on a dual powered server or device is connected to a different PDU. We also make sure that if we have an HA pair of single power supply devices, one device is connected to each PDU. We typically connect the left power supply to the left PDU and the right power supply to the right PDU. In most cases we test the power by turning on all the servers in the rack and intentionally failing each PDU, one at a time. In theory, if either circuit fails, any dual power supply devices will alarm us on a power supply failure, the load will transfer to the other PDU (which will not be overloaded, right?) and your application will not notice. The HA pair of single powered devices will do a failover and failback.

Everything is good, live goes on.

Until a few years go by, a few servers get re-racked, and a few vendor tech's swap a few server parts. Eventually someone will plug something in wrong, and during the next circuit failure you'll have unexpected downtime. (Don't ask me how I know....). The question then is: How do you know that you still have proper power redundancy after the rack has been tweaked around with for a few years?

Red power, Blue power

We bought a couple rolls of colored electrical tape from the local big box home store and wrapped a band around each end of the power cords connected to one PDU in one color and the power cords connected to the other PDU with the other color. (For us, red on the right side, blue on the left.). Now a quick glance at the back of a rack after intrusive maintenance will tell us if we have properly attached dual power supply devices, and more importantly, will tell is that our redundant pairs of single power supply devices are each connected to separate power. Mismatched colors stick out like a sore thumb.

I hate it when I sound like Martha Stewart.

Clueless?

Found this gem on thedailywtf.com.

"...no matter how far software and hardware evolve, there is always someone programming with a sufficient degree of cluelessness and laziness to bring a system to its knees."

 2008-01-30 12:35 • by halber_mensch

Resource Fork

Photos.

Mac's strength. This should be great. Plug in a Compact Flash full of photos. Nothing happens. This is boring. How about automatically asking me to import my CF full of photo's? Nope.

Find 'No Name' on desktop. Open. Navigate to folder with pics. No thumbnails. WTF? I have a flash card full of JPEG's, and the Mac can't even figure out how to thumbnail? Try the menus'. View - nope. File- nope.

I'm driving the worlds best operating system, the operating system that first associated a file type with an application, the operating system that was built to allow you to navigate to a file and use the file resource information to open an appropriate application. A breakthrough in UI design, right? It looks like that operating system, after 23 years of refinement, can't handle a CF card with JPEGs. Doh!

Anyway - I did it the Windows 3.1 way, and opened up an application (iphoto) and used the application to open the files. At least that works.

I find user interfaces fascinating. Some person or group of people, somewhere in California, sit around and try to figure out how I think. That's got to be a tough job.

Updating non-existent software?

I've got a principle, called the 'least bit' principle. The short form of the 'least bit' principle is that when I get a system, I want to configure the system with the least number of bits required for the required functionality of the system.

So for my Mac, the first thing I wanted to do was eliminate all bits (software) that are not required for functionality as I intend to use the system.

So I dragged iTunes into the trash.

It worked. So I emptied the trash.

That didn't work.

A reboot fixed that problem. The iTunes in the trash went away. All is good. I got rid of unnecessary bits.

So now that I have no iTunes, what happens when I run Software Update? You guessed it. Apple offered to update a non-existent iTunes application.
Annoying.

---

We have data

One of our outstanding issues is to figure out how to backup & archive an ever growing mound of data. Right now we perform more or less standard daily/weekly/monthly backups of a hundred-odd servers and about that many databases. Like most shops, we spend tons of time and money on backup servers, tape drives and tapes, all for the inevitable and extremely tense moment when we actually need the backups.

We are crossing a threshold though. Right now, we are near the limit of what we can comfortably slurp across a gigabit network in a reasonable backup window. Of course the option of not storing so much data seems to be off the table. I've never actually seen a user delete unneeded files, or a database magically shrink itself overnight. So for now, expanding the backup system is probably the only short term option.

Over the years we've used a handful of different backup systems and strategies. For our original Netware environment we used Palindrome. Palindrome had a very unique tape rotation strategy, based on a strategy that minimized tape use by never backing up unchanged files that were already backed up on one tape in each of six tape sets. The backups were essentially a continuous set of consolidated and incremental backups added on to the existing fulls. Pretty cool. After a while, after Palindrome disappeared we switched to a combination of Arcserve and Amanda.

As the amount of data grew, we figured that we'd be best off with at single 'Enterprise' backup product, with multi-server backups to tape changers. We shopped around and eventually chose Legato. The rational was that Legato was one of the few vendors to support OpenVMS, and we knew that we wanted all backups to use the same system, so we jumped in with both feet into the Legato pool. We initially chose a Windows 2003 server for the backup server, with fiber channel attached tape libraries. That system, Legato on Windows with fiber channel attached tape libraries, worked for quite a while. But the data kept growing, and as the data grew, so did the headaches, primarily around tape drive reliability. They simply weren't reliable. We were spinning the drives about 8 hours per weekday, and about 36 hours per weekend. They didn't like the abuse, and we got tired of coming in weekends to fix them.

Another trip to the vendor, a handful of meetings, and a big purchase order got us a major upgrade. We now dedicate a Sun T2000 to the backup server role, and we've set up huge disk pool and both virtual and real tape libraries. This system is really only a couple years old, and has only been fully implemented as designed for less than a year. We built a system that allows for fast backups, fast restores, multiple copies of backed up data and quite a bit of flexibility.

The design goals were:

1. No dependency on tape for the initial backup. (too unreliable)
2. Backups off site as soon as possible. (disaster recovery)
3. A fully recoverable set of data on disk, not tape. (fast recovery)
4. Duplicate copies of each backup, on different media. (paranoia)
   
5. An archive to tape, for normal retention.
6. No headaches on weekends
   

Here's how it works:

Slurp the data:

We slurp up all the data using standard Legato agents and client side compression to a big disk pool. Right now we have a 10TB ZFS file system made up of five 2TB luns on an HP EVA8000, all mounted on a Sun T2000. The file system is configured as a single ZFS disk pool, striped and carved up into several ZFS file systems. We don't need ZFS to provide redundancy on the Lun's, because the EVA provides disk redundancy for the entire disk pool. The T2000 catches the data slung across the network from the servers across the network and writes it to disk.

That part has worked extremely well. The T2000 (an early, slower model) can read & write at near gigabit speed across the network while reading and writing to the disk pool at better that gigabit speeds. It does I/O real well.

Migrate the data:

As soon as the data from a server is spooled to disk we start the process of copying the data offsite and archiving it to tape. The offsite copy is stored on a fiber channel connected virtual tape library (HP's VLS) located about 10 miles away in our secondary datacenter. The archive copy is written to conventional tape. Unfortunately Legato doesn't have software that that can migrate data from the disk backup to tape and still retain the copy on disk (fast recovery). They insist on deleting the disk pool data as soon as it is on tape. So we wrote our own data migration daemon, (in Perl, of course).

The migration daemon watches the disk pool, and as soon as it sees a backup job (save set) finish writing to the disk pool, it starts cloning the job to both the off-site VLS and the real tape library. It also watches the free space on the 10TB disk pool and purges old save sets off the pool as the pool fills with new save sets, but only if the save set is already cloned to both the VLS and tape.

So as things are today, we always have a fully recoverable set of data on the disk pool (fast recovery), the VLS (off site), and conventional tape (archive).

But the data keeps growing.

We've been watching our backup windows grow, and as they grow, we get closer to the point where we have to re-think the whole mess. Right now, we have close to 12TB of data that gets backed up per full backup. That gets compressed down to around 6TB of data when it hits the disk pool, the VLS and the tape library. Moving that much data around is getting to be tough to do, without going back to the whiteboard.

I did some quick math based on estimates of network throughput and volumes of data. We'll likely have to re-think the network based backups, or at least re-think how we use the network for backups. We know that we can use the fiber-channel more, and we know that we can build a dedicated backup network, but both of those options have downsides, either in complexity or security. We also know that there are smarter alternatives to conventional daily/weekly/monthly full/incremental backups available from a variety of vendors, but making that large of a change introduces lots of work effort and risk.

We are still thinking about this one.

A MacIntosh

I used a Mac once. A notebook. A long time ago.

I don't remember the year or the exact model, but I remember it went 'QUACK' whenever I did something wrong, and it had dead pixels. It might have been a '140' or something like that. You could display any color you wanted on its display, as long as it was black or white.

I really liked a couple things about it. It was great at converting documents. I used to have to convert documents between WordPerfect and Word. The Mac Word version was better at converting PC WordPerfect docs to & from PC Word than either of the same programs on the PC.

I quit using Mac's sometime around the time that computer went away. There wasn't anything about the computer that I thought was interesting, and it certainly couldn't replace my OS/2 desktops, so for me, Mac's dropped off my radar.

Now they are back. I finally jumped in to the Mac pool with a borrowed PowerBook. I'm finding it interesting, amusing, and sometimes frustrating. I'll keep notes.

Tethered

I'm not a big fan of notebook/laptop computing. 

I'm just not excited about dragging around 10 pounds of stuff just so I can check my mail & edit a document, especially when with 20lbs of stuff, I can live in a foreign country for a week. Having a laptop, in my opinion, doesn't really un-tether me from anything. I still need to be within 9ft of a power source at least once every couple hours, I still need to have a flat table or desk (or live with scorch marks on my polyester leisure suit pants), and I still need a posture correct chair. Using a 'laptop' as a lap top computer didn't work very good with my Epson 386-16, and after trying a few more times recently, it still doesn't work. I either have to live with 1-2 hours of untethering at a time, or I have to live within range of power, and I have to drag power brick around, or an extra battery, but I can't swap batteries without shutdown and reboot. And even worse, if I dose off while relaxing watching movies or reading PDF magazines or forums, the 'laptop' tends to crash to the ground and wake me up.

And of course the elephant in the closet is the simple problem of backup and recovery. From what I see, the vast majority of users with laptops seem to think that they will never crash a hard drive, and will never get lost or stolen. Guess what - they will.

The 'Laptop' needs to die. Let's move on.

The concept of bringing my computing environment with me wherever I go is pretty much an '80's concept made necessary by the lack of wide-spread networking last century. Things are different now. The network works. I don't need to drag my entire computing environment around with me - I only need to bring along enough resources to be able to access my stuff from whatever network happens to be available. That is how I've operated since about 1992 or so when I first figured out what a pain it is to drag a laptop down the street, and that is why I've used VNC, RDP, Citrix, SunRays and Palms. A laptop with extra battery, power brick, mouse, mouse pad, etc, is heavy enough that 'road warriors' have roll-around laptop cases. You've got to be kidding. The few times that I've carried a 'laptop' around in the last 15 years it was strictly to use as a device to access a real computer somewhere (like via Citrix).

I don't want a computer in my pocket - I want an access device that lets me access a variety of computing resources and applications. I want it to fit in my pocket, weigh a few ounces, and last a day without having to tether to a power source. I want my data to be stored someplace secure, available and accessible. An obviously I want it backed up.

So what options are out there? It is, after all, the 21st century, and some companies are starting to get it. We have a reasonable selection of UMPC's that although they are still a complete computer, they have way more portability, and the flash based ones have way less heat output and much longer battery life. One could actually imagine holding one of them for hours at a time while reading e-magazines or news. The UMPC's can be equipped with EVDO, UMTS and WiFi, so the probability of having network access is pretty good, and they can function as 'access devices' instead of full computers. Unfortunately flash based UMPC's cost way too much for an access device, and so far they are stuck with using an OS that is made for 40GB hard drives & 1GB RAM.

Another UMPC sort of device is the N800 from Nokia. Is has the size, weight of a large phone, but has great 800x480 video and can work with as little as a couple hundred meg of RAM and flash. The battery life is 8 or so hours, or maybe all day, depending on how hard you surf. It can hook up to most any WiFi network and can grab internet access from your phone via bluetooth DUN when WiFi is not available. It runs a bunch of IM protocols, an RSS reader, a media player and a decent browser. Application support is lacking a bit though, with the open source crowd working to fill the gaps. It is not a PDA and can't edit a document, but it can be hacked to run an RDP thin client. The browser works - I use it more than any other browser.

And of course Sun sort of has the concept with the latest Sun Ray. The newest notebook-like Sun Ray clones can grab a network, connect up to your computing resources and act as a remote display/keyboard. Sort of like carrying around a Citrix client instead of a computer that can load a Citrix client. Purely an access device. If there is no access, there is no computing. We've been using the desktop Sun Rays for years. They work.

The smartphone crowd keeps trying to build an 'access device' that is primarily a phone, but can do enough real work to break the 'laptop' dependency. Palm came close for a while. I had a palm phone that could run RDP & SSH for remote access, could check or sync my calendar and mail, and could access the web 'good enough' to keep me from having to carry a 'laptop' while traveling. Some of the Symbian and Windows based smartphones look like they come pretty close also, but they still assume that I am willing to tether myself to another computer or laptop periodically to sync my data and backup my phone. Sorry - I don't want to be tied to a laptop. They are dead, remember?

What about the iPhone?

If I don't mention the iPhone, nobody will read this. So here it is. The iPhone might be a step in the right direction. Forget the eye-candy. I don't care if my notepad looks like it is tearing off sheets or turning pages. That sort of silliness serves only to shorten my battery life. What I care about is Apple's software development rules for the phone - at least as far as they look right now. It looks like Apple is pushing towards hosted applications that are accessed via the network, with the phone's browser as the access device. Apple is assuming that the network to access the applications will always be available. I like that. It is that concept that is most interesting. The pundits that are complaining about not being able to install applications are going down the wrong path. The applications and your data are already on the network, in a data center, secure, and backed up. You don't need apps on your phone. You need access to your apps from your phone.

The down side? An iPhone is still tethered to a computer. You can't activate it or do anything interesting (like download a PDF) without using iTunes. That is simply not acceptable, and for that reason I am not interested in an iPhone. I want to replace the big, power-hungry home computer, not tether another USB device to it. The iPhone needs to be able to stand on its own two feet. When it grows up, it will. No doubt.

What else is new?

I ran across an interesting concept in desktop computing from Zonbu. $250 for the computer, or $99 + $13/mo for the computer plus on-line storage, support and updates. It has a flash disk, Linux, and what might be a usable set of applications.

Here's what is interesting:

The computer uses 13 watts, no fan. If you leave your computer on all the time, it pays for itself in a year compared to an ordinary desktop. This sort of power consumption is down in the range of a Sun Ray. The lack of a fan means that it is quiet - even quieter than a laptop. So it can stay on all the time. And even more interesting is the default storage. The device uses S3 to store its data. This makes for an interesting model. At home, where you might want a real screen & keyboard, you'd use this appliance-like device to access your documents via the network and S3. While away, an S3 client on a smartphone, PDA, iPhone, N800 or whatever happens to be around gets you access to your data. It is not hard to imagine a 1lb 1024x800 WiFi enabled clamshell version of this as a laptop alternative. And the best part is that I could replace my redundant home network storage devices and even my iBackup.com storage with always-on S3. It is a thicker client than a Sun Ray, but thinner than a PC. If it used way less power than an ordinary laptop, the battery life could be long enough and my pants wouldn't melt & stick to my legs. Perhaps it is sweet spot?

The guys that are clinging to their PC's are like the VT100 terminal guys from 25 years ago. They thought that if they made their terminals do fancy graphics, they'd stave off the PC onslaught. They were wrong. The world wanted freedom from centralized computing. But now, 25 years later, with my applications and data scattered across three phones, two laptops, a desktop, a handful of servers & network storage devices both at home and at work, a Nokia N800 and who knows what and where else, the idea of centralized computing resources accessible from a variety of devices looks pretty attractive.

Long live the iPhone. Jobs might steer us in a useful direction.