[Tfug] OT: Disk testing

[Ronald Sutherland]

I don't have much time during week days.. between a Pima class and work 
I can't even do proper lurking. But some fast comments, current limit .. 
ya .. its no fun to plug into a hot source with high current limit (pop 
.. ekk .. wait for sight to return... a good day). So a number of 
smaller independent sources with lower current limit is much better. At 
any rate I don't think power usage should be a concern, the real issue 
is elegance or lack of, and big globs of wires lose in that competition 
every day. Even so...

Not sure if you were going to do the testing or someone else, or how 
many units needed tested. What I have in my mind is a PC (or more as 
needed) with 2-5 mobile docks. The operator installs the HD's in dock, 
plugs them in PC and turns on the power. The test runs automatically 
after boot and unmounts all HD's at end of test, leaving PASS/FAIL info 
on screen and log (or stuffs it into a network server). The test 
operator then powers off the PC and removes HD's. If a surface scan is 
anything like mirroring then that takes me a long time on >60GB HD's.

The test PC(s) ATX supply could be smaller (200-350W) if fewer HD's are 
used, that keeps a reasonable current limit and is cheep, also no cards 
or extra boards needed (just mobile docks). Don't know about you but I 
have a few extra PC's at work (and home), putting them (the work ones) 
to use like that would be my path of least resistance, although I would 
expect them to self destruct from power cycling. I'm not sure power 
cycling USB hardware is any better but it may be, most of this consumer 
crap is full of low end electrolytic capacitors and is not really tested 
for reliability.

The fixture is a mid tower PC case with 2 to 5 5-1/4" mobile docks 
mounted. You don't have to mount the HD's in the dock carrier, just plug 
in the IDE cable and power, set the HD down in carrier, and slide the 
carrier into the case. Everything is enclosed and out of the way, no 
wires dangling on bench, to drop a screw driver on.

Bexley Hall wrote:
> Hi, Ronald,
>
> --- Ronald Sutherland <rsutherland at epccs.com> wrote:
>
>   
>> You may be right on most accounts, testing HD's is
>> not the sort of thing I do,
>>     
>
> Nor I!  :>  I design embedded systems -- hence the
> inherent bias to look for *that* form of solution  :>
> ("when what you have is a hammer, everything looks
> like a nail")
>
>   
>> I do get payed to develop test for power
>> conversion products, but its not the same.
>>     
>
> Yup.  I'd imagine you have to worry more about
> environmental stresses, etc.
>
>   
>> In general I don't like to invent, its usually a
>> wast of my time, and it always gets undone in a 
>> few years when the setup has to go to Mexico or
>> China.
>>     
>
> Nor do I.  I *reuse* (I've been doing this sort of
> thing for a fair number of years so I have LOTS of
> already debugged designs -- software + hardware -- to
> cull from).
>
>   
>> I like things with a well defined interface, 
>> both hardware and software,
>>     
>
> Hence the beauty of the IDE connector.  Or, at
> the other end, the USB connector.
>
>   
>> I've even made up my own term "Modular 
>> Components Off The Shelf". Its Maintained, invented
>> and extended by 
>> not-me, but it should be open if needed. If the ATA
>> interface is simple, 
>> thats all the more reason to use it. Adding layers
>> like USB or Ethernet, 
>> have caused me some problems with other data buses,
>> and I have limited 
>> kernel or driver software skills, although I have
>> done a lot with 8 bit 
>> MCU's without a proper OS. So the idea of an MCU
>> sounds good, but 
>>     
>
> Problem with a small/cheap MCU is tying that to
> a host gets "expensive" (in terms of software
> if you want any sort of bandwidth).
>
>   
>> basically that is what CORAID is doing with ATA over
>> Ethernet, which would also seem elegant.
>>
>> http://www.coraid.com/
>>
>> The problem with the Coraid product is that it will
>> not stand up to much 
>>     
>
> It also looks to have a higher layer of abstraction
> than is needed.  E.g., conceivably, you could
> exchange commands with a "dumb" remote like:
> "report drive geometry" (assume the remote adopts
> that geometry unconditionally), "perform random
> read/write/verify test", "scan surface", "scan
> surface with pattern", "report results".
>
>   
>> abuse. If speed is not an issue, then I have no idea
>> what "comprehensive 
>> surface analysis" is, anyway USB is clean at least
>> for wiring, however a 
>> mobile rack type thing can help when you are dealing
>> with a lot of parallel control lines.
>>     
>
> But you don't have a lot of parallel control lines
> in this configuration -- unless you run the IDE bus
> up the back and add a set of buffers at each
> drive (to isolte the drive during [un]plugging).
>
>   
>> The wiring mess is in an enclosed fixture and 
>> the device under test is put in a carrier, then
>> loaded into the fixture, 
>> its a basic method used to test many products.
>>     
>
> Yes, but that's a lot of effort into building
> a fixture.  I'm looking for something "on the cheap"
>
>   
>> PC's do eat a lot of space, and thats a real 
>> problem, As far as power, a PC is less wasteful 
>> over all.
>>     
>
> That assumes you do it with a *single* PC,
> single monitor/keyboard, etc.  And, that the PC
> only has what is needed -- no extra peripherals
> that don't contribute to the task of checking
> the disk.
>
> If you start using multiple PC's -- and multiple
> monitors, etc. -- then each new PC adds another
> big chunk to the effective "quiescent power" :>
>
>   
>> Every power conversion device eats some bias power
>> (2% or more) and smaller supply's are the most, 
>> with high bias (~10%) and are less efficient.
>> So a single PC supply (say 600W) running everything
>> is about the most efficient method power wise. 
>>     
>
> But that doesn't mean that you *still* can't use
> a single supply to "power everything" -- even if it
> is NOT in a PC!  E.g., you can still run pigtails
> to each of the "external drives" from that single
> supply (assuming you watch IR losses)
>
>   
>> Many PC's have a power supply that is a real POS
>> and need replaced with a good one that includes
>> PFC. The power supply will only use what is
>> needed, plus its bias power, and efficiency factor.
>>     
>
> The other downside to a single supply is that it is
> larger (capacity) and thus overloads at a much
> higher load (e.g., now you have to worry about
> someone plugging/unplugging a 5VDC line that may
> current limit at *50*A instead of 10).  And,
> anything that goes wrong (e.g. someone failing
> to switch power to the individual drive *off*
> before unplugging/plugging and "misconnecting"
> to the load -- or, two power pins touching
> a piece of metal, etc.) now you put all of the
> devices on that load at risk -- i.e. you have
> to retest them all, reboot the PC, etc.
>
> But, the single PC issue is still a bottleneck,
> IMO.  IDE cables are supposed to be *short*.
> So, you end up stuffing a lot of IDE cards into
> a machine and then *hoping* you can fan those
> out well enough to connect to N drives mounted
> outside the PC, etc.  It just seems like a
> cable dressing nightmare.
>
> I think the quick-and-dirty solution would be to
> find some very *small* PC motherboards with an
> on-board NIC and IDE -- one or *two* ports.
> Then, boot them diskless (eliminates the need
> for a "master" disk on the primary IDE port)
> and let each handle two drives -- one on primary,
> the other on secondary IDE port.
>
> The "PC" could be reasonably slow -- since all it has
> to do is move bytes to/from the IDE.  And, you could
> use the NIC to talk to the host (much easier than
> using USB).
>
> The key would be to find something very *small*
> so that you could almost slip it under the UUT
> without taking up much space, etc.  If everything
> is on the motherboard, then it could be a low
> profile since no ISA/PCI cards sticking up...
>
>
> __________________________________________________
> Do You Yahoo!?
> Tired of spam?  Yahoo! Mail has the best spam protection around 
> http://mail.yahoo.com 
>
> _______________________________________________
> Tucson Free Unix Group - tfug at tfug.org
> Subscription Options:
> http://www.tfug.org/mailman/listinfo/tfug_tfug.org
>
>