Parallelism in disk readout?

Suppose you have a high performance RAID array; just for laughs say RAID6+1. The read speed is significantly faster than the read speed for a single disk, and the mirroring helps preserve the data from loss when one drive dies. Seems cool, though pricy; and people use them all the time.

But how easy is it to back up? Don't ask. Unless you get the full bandwidth it takes a long time to back up 80TB or so, and even with full bandwidth it takes a while.

Suppose your chassis was designed for a particular RAID configuration (instead of being customer-selectable). Let it be, instead of Nx2, Nx3. Instead of having a single mirror, you have 2, one of which is read/write and the other of which is "write only" for the main controller.

However, each of the drives in that third "write-only" row have their own independent controller. (For the moment forget about spares. They add some complexity but nothing overwhelming.) Each controller is linked to a multiplexer that manages 2 or 3 drives and communicates over a fibre-channel connection.

Now suppose you have a second chassis with N drives, each of which has its own controller similarly linked to a multiplexer. A bundle of fibre-channel cables connects one chassis to the other.

1. The chassis is told that a backup is at hand, and senses the presence and health of the other chassis
   
2. The chassis pauses all pending reads and writes
   
3. When all active reads and writes have completed, the controller takes a snapshot, and releases the pending reads and writes
   
4. The controller releases all drives in the third (write-only) row and instructs the sub-controllers to begin
   
5. Each sub-controller communicates with its companion in the backup chassis, and writes the blocks on its disk to the other.
   
6. When the last copy is done, the sub-controllers relinquish their disks back to the main controller.
   
7. When the main controller has full control again, it pauses pending reads and writes again.
   
8. When all active reads and writes have completed, the controller takes another snapshot, and releases the pending reads and writes
   
9. The controller uses the incremental change between snapshots to update the disks in the third row.
   
10. When the update is complete, another pause and snapshot is required to take into account the changes that happened while the update was happening
    
11. A few iterations may be needed until the incremental is small enough that the system may be acceptably frozen long enough to make the final update.
    
12. The three rows are now again in sync, and the backup chassis holds a consistent backup of the system. It can be removed and stored, or whatever is desired.
    

Restoration uses the backup chassis as a source and a high-performance one as a target.

How long will such a backup take? A 3TB disk being read at 250GB/hour will take about 6 hours. That's better than a full backup any other way I know of (I'm not counting synthetic full backups. If you lose one of the "deltas" you're up the creek.)

I shudder to think how much this would cost. But I wish we had a few at work.

Which type of stroke?

The blood supply to the brain is a web encased in a rigid box. I'd not expect the blood vessels to be able to expand much--between brain tissue and other fluid around them, there's not much to compress and no room to go thanks to the skull. But there should be a little elasticity to the blood vessels, and there are so many that the effect should add up.

With each heart beat we have a more or less predictable pressure wave going into the network, and so there should be a consistent pressure wave coming out. It can't be exactly the same shape, because in different parts of the network blood travels different distances facing different impedances, but it should be consistent.

Now suppose some major vessel has broken and there is bleeding in the brain. If the blood loss is significant I'd expect there to be a change in the wave shape. In particular, blood would leak out of the broken vessel and slowly reenter, giving the output pulse a longer tail.

So the obvious questions are:

1. Are the output pulse shapes (maybe convoluted with the input pulse if need be) consistent enough between people that we can use differences for diagnosis?
   
2. Is the tail from a leaky vessel discernible in the output pulse, and if so at what level?
   
3. Are there other things that would mimic this?
   

You have to measure the input also, since a long tail on the input is quite possible, and you'd need to correct for that.

If it is possible to distinguish a leaky vessel-brain from a normal one, it would be possible to distinguish a broken blood vessel stroke from a clot blockage stroke. (The instrumentation shouldn't be very expensive--transducers and some ADCs and a small processor to compare the results--a display if you want to get some human input.) Since you don't want to use blood thinners to treat someone with a broken blood vessel stroke, and since getting treatment fast is vital, this might be useful in an ER.

I don't think it would be possible to positively ID a blockage this way, since that would merely appear as larger impedance, and without some prior measurements you wouldn't know if this was normal or not--brains are different.

Font Directories

Eldest Son told me that the printing programs he is learning lock out most fonts so that the user isn't swamped with pages upon pages of fonts. The "book" classifies them as text, dingbats, symbols, script, and a few other things.

That doesn't make any sense. I don't know enough about fonts to devise a scheme myself, but just from inspection you can make general categories. My untutored eye says there are blocky fonts, and shaded fonts, and fonts that are good for tiny print, and such like things. A given font may wind up in several different categories--which seems harmless.

Suppose one imposes a directory structure on the fonts, and uses links to insert them in multiple places as needed. For example, if you want to search alphabetically, there's a "B" directory that contains links to all the fonts with names that start with "B". If you want German Black Letter, there's a directory for German fonts, as a subdirectory of the NonEnglish directory. The icon for the directory has not one (as Windows uses now) but multiple images of the letter A illustrating some of the directory's contents: or from subdirectories if the directory contains nothing but subdirectories.

It'd take some work to devise the scheme and implement it, but searching that would be a lot easier than scrolling down a few hundred options.

Aspergers and theater revisited

A few years ago I mused about the possibility of teaching roles and how to detect them using theater. We were not able to schedule any kind of formal therapy along these lines. There was a local person who was interested in this sort of thing, but unfortunately she/he had some personal issues and went transsexual--and, though she had been given no introduction or description, our daughter was weirded out just meeting him/her/whatever. That did not seem a good match. It was encouraging to think that even with Aspergers our daughter could instantly detect that something was seriously different, and probably wrong. But that group wasn't going to work for her. We encouraged our youngest daughter to join a theater group in high school, partly to help her with finding a role in the chaotic mess and partly to help with learning about role-playing and detecting roles as described earlier. She found it frustrating at times, but learned to enjoy it. It is nice to have clear-cut lines and roles, and she did well at that. She loves opera and the great romantic/tragic characters, but opera doesn't always have the most realistic dialog and roles... I wish I could say there were breakthroughs, but this has been and still is an incremental process.

DJs for children's play

The athletic club's family new years' night had one gym full of "bouncy" toys, with the usual castles and cage of plastic balls to flounder in--and a DJ playing amusing songs at full volume.

But what do the kids really like? Do they like loud music, or soft, or something else entirely?

This is a job for the psychology department: a student project!

Every week for a couple of semesters hold a kid's fun Saturday afternoon in a gym. Put the same play equipment on both sides, in the same positions. In the middle put chairs for the parents (who are told what the study is about, and asked not to lead the kids one way or another) facing in, and a large "hanging strands" curtain in between. The "curtain" will be fun for kids to run back and forth through, which will randomize the side they eventually wind up on.

On one side put a DJ (one of the psych students) with a sound system angled to put the loudest sound on one side of the room. In other words, one side of the room has the music (or whatever) and the other has much less music volume.

Every ten minutes count the number of kids playing on the "bouncy" equipment on the music and the "quiet" side of the room. Repeat this for several Saturdays, and then change the type of music.

We can't put two kinds of music in the same room without dissonance, and can't separate the rooms because the kids will want to be with their parents. Therefore we can only study degree of loudness preference for various types of music.

Options would include

• Loud rock
• Quieter rock
• Loud pop
• Quieter pop
• Loud "kid's music"
• Quieter "kid's music"
• Playground noise (kids playing)
• Lullaby music

If the gym holds about 60 kids at a time and you run the experiment for 5 Saturdays (and we ignore the fact that some kids will come repeatedly) we get 300 measurements. You'd expect some fraction to be with their parents or running through the curtain--so say 200 are in the bouncy play areas. An even split would be 100 on each side (if they have no preference), so you should be able to measure preferences at the 10% level or better.

I'm not sure how much this would cost: gym space plus equipment rental plus fabricating the curtain plus music fees...

Multi-touch screen sound mixers

I think that Band in a Box is missing something, if I've read their documentation correctly. They provide all sorts of sounds, but not mixer consoles. If you create a set of instrument sounds, you will want to adjust their relative volume, probably on a dynamic basis, and the obvious way to do that is with some variety of multi-touch screen interface to the computer. You could drag-and-drop to assemble and connect the board to the instrument sources you've created, and use virtual sliders to adjust their relative volumes--recording that information as well for later playback. I assume that frequing the signal is done at the instrument level and not the mixer board level, which simplifies the board layout somewhat.

It wouldn't be as easy to use as a real mixer without touch feedback, but the possibility of easily recording and later revising the changes would seem to make up for it for the novice without the bucks for a full system--which is one of Band in a Box's markets.

Scent of a human

Has anybody tried to get a thorough collection of all the chemicals we exude/exhale? We all know about CO2 and H2O, and methane; and some of us know of a few others: chemicals from the bacteria in the armpits, and so on.

Suppose you took clean (or at least thoroughly analysed) O2/N2/CO2/H2O, chilling it down to nearly liqufaction to get rid of impurities and then heating it back to room temperature. This is the input air for a sealed (baked out) bare room with a recently bathed naked man standing in it. The output air is also chilled to condense out the water and then super-chilled to condense out the other chemicals he has exhaled or exuded from his skin, letting only the O2/N2/CO2 escape.

We already know about the relative rates of exalation of water and CO2. What else is in that puddle we'd condense out of the used air? (Trying to understand the shed skin cells is probably too huge a task.)

I don't know how good we are at "figure out what's in here" chemical searches; though we can figure out ways of detecting known chemicals very well. But there's bound to be a lot of different chemicals present. They may vary by time of day; they almost certainly vary from men to women, and there's a hint that they may very with a woman's menstrual cycle.

Something's there. I wonder what it is? Too bad dogs can't talk.

UPDATE Some people look for things like benzene or chloroform in the breath to measure uptake of contaminants, but these are specific searches. I'm thinking of more basic research. A systematic search using many subjects in many states of health might be useful in finding new diagnostic techniques. How much does your body chemistry change when you get such-and-such a disease? Can you "smell" the difference? Once you know what to look for, it might be easy.

Also, chromatography is a very powerful tool, and can separate out chemicals in several different ways.