Re: [9fans] Anatomy of a vblade image

Re: [9fans] Anatomy of a vblade image

[matt]

(In the opposite style to usual, this was meant to go to the list, not
private :)


> i didn't want to sound over-confident.  but vblade
> is very simple; i don't see how it could have data-loosing
> bugs.
I can't even remember how I made it blue-screen but I managed it a
couple of times. I think one time was returning the string "hello" to
whatever read requests were made :>

> what are you doing with (to) vblade?
>
Trying to do something interesting enough to write a paper for IWP,
I'll get into detail when it works

> snoopy's your friend.  you can test just by killing vblade
> and watching the packets go by.
>
Wireshark's pretty good too. It even knows how to decode 9p streams.

> good deal.  about a year after i wrote vblade, i used it
> for a couple of months while i saved up for a stand-alone
> aoe target.
:> I'm hoping to get a decent bit of kit together for the office here. I
already boot the Linux terminals from AOE, hope to soon add the Windows
machine.
I built a Plan 9 installer iso that includes aoe & fs as a target but I
haven't managed to get a working Plan 9 system out of it yet - I got a
bit side-tracked :)

In the Linux world kvlade looks the way to go for performance but it
didn't work on AMD64 when I tried it. I guess the Coraid boxes make Plan
9 go fast enough so I just need to stop worrying about it.

m

Re: [9fans] Anatomy of a vblade image

[erik quanstrom]

> In the Linux world kvlade looks the way to go for performance but it
> didn't work on AMD64 when I tried it. I guess the Coraid boxes make Plan
> 9 go fast enough so I just need to stop worrying about it.

plan 9 is a good platform; it is very speedy and doesn't do anything
to you.  we didn't need to make it go fast.

- erik

Re: [9fans] Anatomy of a vblade image

[John Waters]

Indeed,
I have been toying around with the idea of building a larger scale
version of Mr. Minnich's lunchbox using Mini-ITX boards..
But I have no justification, since virtually everything i need to do
with plan 9 happens impossibly fast on just a simple vmware instance.
Plan9 is sufficiently speedy to stifle my desire to build a fast system.

John

On Tue, Jul 15, 2008 at 2:49 PM, erik quanstrom <quanstro@quanstro.net> wrote:
>> In the Linux world kvlade looks the way to go for performance but it
>> didn't work on AMD64 when I tried it. I guess the Coraid boxes make Plan
>> 9 go fast enough so I just need to stop worrying about it.
>
> plan 9 is a good platform; it is very speedy and doesn't do anything
> to you.  we didn't need to make it go fast.
>
> - erik
>
>
>

Re: [9fans] Anatomy of a vblade image

[erik quanstrom]

> Indeed,
> I have been toying around with the idea of building a larger scale
> version of Mr. Minnich's lunchbox using Mini-ITX boards..
> But I have no justification, since virtually everything i need to do
> with plan 9 happens impossibly fast on just a simple vmware instance.
> Plan9 is sufficiently speedy to stifle my desire to build a fast system.

small pcs price/performance ratio is too high for me if all i want is a pc.
it's hard to beat a $50 motherboard and a $34 processor.

but for other small appliance-y things, some very-low power mips or arm
boards are attactive.

- erik

Re: [9fans] Anatomy of a vblade image

[matt]

>
> however you can use the -i switch to initialize a blank
> file.  (you're free to use a raw disk or partition, too.)
>
> - erik
>
>
or say "damn, where did my partition table go !"

I can't remember if I lost any work :)

strings /dev/sdC0/plan9 will let me know eventually !

note to self, be more careful when messing around at storage :>


I'm serving drives to WinXP64 via AOE with http://winaoe.org/
I've had a few blue screens but I think they have been caused by the
server side serving bad data. If the vblade goes away after being
mounted the NT kernel keeps trying to read it which amounts to waiting
for network packets to arrive and of course they never do so any NT
syscalls take forever (though they run eventually). It does come back to
life even if you vblade a different sized disk.

Re: [9fans] Anatomy of a vblade image

[erik quanstrom]

> I'm serving drives to WinXP64 via AOE with http://winaoe.org/
> I've had a few blue screens but I think they have been caused by the
> server side serving bad data.

what do you mean by "server side"?

> If the vblade goes away after being
> mounted the NT kernel keeps trying to read it which amounts to waiting
> for network packets to arrive and of course they never do so any NT
> syscalls take forever (though they run eventually). It does come back to
> life even if you vblade a different sized disk.

it is expected behavior for aoe devices to be
able to take short vacations.  our main fs uses
aoe.  we can upgrade the shelf without bringing
down the fs.

it does sounds like the free windows aoe driver
needs to handle the case where storage goes
away more gracefully.  it may be that it is retransmitting
too agressively.

- erik

Re: [9fans] Anatomy of a vblade image

[matt]

erik quanstrom wrote:
>> I'm serving drives to WinXP64 via AOE with http://winaoe.org/
>> I've had a few blue screens but I think they have been caused by the
>> server side serving bad data.
>>
>
> what do you mean by "server side"?
>
My dodgy programming :> I'm hacking around in vblade.c
Sorry, I see now I should have made that clear.


>
> it does sounds like the free windows aoe driver
> needs to handle the case where storage goes
> away more gracefully.  it may be that it is retransmitting
> too agressively.
>
>
I haven't had the ethernet sniffer looking at it


I got my data back btw. I ran the installer and guessed at what I'd told
fdisk :>

Re: [9fans] Anatomy of a vblade image

[erik quanstrom]

> I shuold just mail quanstro I guess but maybe someone made a tool
>
> /n/sources/contrib/quanstro/src/vblade
>
> The vblades served by this particular vblade require some magic in the
> first few bytes.
> Vblade on other platforms will quite happily serve a bunch of zeros
>
> Anybody got a script for making an empty vlbade image for serving with
> the above vblade ?
>
> Save me making one :)

plan 9 vblade is a clean room implementation.  friday
evening entertainment. :-)

so i use a header which has magic, size, configuration
string, and the shelf and slot of the vblade.  vblade
checks this magic on startup to determine these parameters.
however you can use the -i switch to initialize a blank
file.  (you're free to use a raw disk or partition, too.)

- erik

[9fans] Anatomy of a vblade image

[matt]

I shuold just mail quanstro I guess but maybe someone made a tool

/n/sources/contrib/quanstro/src/vblade

The vblades served by this particular vblade require some magic in the
first few bytes.
Vblade on other platforms will quite happily serve a bunch of zeros

Anybody got a script for making an empty vlbade image for serving with
the above vblade ?

Save me making one :)

Matt