[saga-rg] proposal for extended file IO
Andre Merzky
andre at merzky.net
Mon Jun 13 13:40:08 CDT 2005
Hi John,
Quoting [John Shalf] (Jun 13 2005):
>
> Hi Andre,
> I think there is a 4th possibility. If each of the I/O operations can
> be requested asynchronously, then you can get the same net effect as
> the ERET/ESTO functionality of the GridFTP.
I disagree. You can hide latency to some extend, but your
throughput suffers, utterly.
Imagine Jons use case (its a worst case scenario really): You
have a remote HDF5 file, and want a hyperslab. Really worst
case is you want every second data item.
Now, if you rely on read as is, you have to send one read
request for every single data item you want to read. If you
interleave them asynchroneously, you get reaonable latency,
but your throughput is, well, close to zero.
If you want to optimize your buffersize, you have to read
more than one data item CONSECUTIVELY. Since the use case
says you are interested in every second data item, you
effectively have to read ALL data.
Same holds if you want every 10th data item - only the ratio
gets even worse.
So, interleaving works only efficently for sufficiently
_large_ independent read request (then its perfect of
course).
> The advantage of simply
> embedding that functionality into the higher-level concept of
> asynchronous calls is that if the underlying library does *not* support
> the async operations (or some subset of the operations cannot be
> performed asynchronously) , you can always perform the operations
> synchronously and still be able present
>
> I do not like plan A or B for the reasons you state. I do not like
> Plan C because it is too tightly tied to a specific data transfer
> system implementation. I would propose a Plan D that simply augments
> the Task interface of SAGA. For example, if you allowed the user to
> fire off a number of async read operations
> Task handle1= channel.read();
> Task handle2=channel.read();
> container.addTask(handle1);
> container.addTask(handle2);
> container.waitAll();
>
> The read operations in this example can be submitted as an eRead
No, it cant efficiently expressed as an eRead anymore. The
implementation sees only a number of reads, but won't be
able to recognize any usable pattern (or at least that
REALLY tough). So, the implementation cannot send the
request
filename hyperslab=([2,3,4][3,4,5])
but has to send each read command
filename read(2,1),read(4,1),read(6,1) ...
Worst case: you send more than one byte as command for every
single byte you request as data (Ugh!).
(BTW: Think HDF5 file driver: that is exactly the problem
there: the file driver does not know about semantics
anymore, but sees only read, write and seek. Hence its
difficult to efficiently implement a remote HDF5 file
driver. Andrei did that, but we had to smuggle semantic
information down to the IO level...)
I think there is NO data model agnostic way which provides
good efficiency for remote data access. I'd be happily
convinced otherwise, but thats what I see right now.
If (only if, again: please convince me otherwise), so IF one
accepts that data model info have to be part of a remote
read request, an eRead thingle a la GridFTP is the best
(most generic and most simple (!)) solution I know of.
> operation or they can be in separate threads, or they can simply be
> executed synchronously when you call waitAll() (this is in fact how
> some of the async MPI I/O was done on the first SGI origin machines...
> it was meant to look asynchronous, but in fact the calls did not
> initiate until you did a "Wait" for them).
>
> Anyways, using the task interface provide more degrees of freedom for
> implementing async I/O than simply supporting the GridFTP way of doing
> things and it meshes gracefully with I/O implementations that do *not*
> offer an underlying async execution model.
I think the task model and the proposed eRead model are
orthogonal. The task model provides you asynchroneousity,
the eRead provides you efficiency (throughput).
Also, as a side note: I know about some of the dicussions
the GridFTP folx had about efficient remote file IO. They
have been similar to this one, and the ERET/ESTO model was
the finally agreed on.
Cheers, Andre.
> The only modification that would be useful to add to the tasking
> interface is a notion of "readFrom()" and "writeTo()" which allows you
> to specify the file offset together with the read. Otherwise, the
> statefulness of the read() call would make the entire "task" interface
> useless with respect to file I/O.
>
> -john
>
> On Jun 12, 2005, at 11:02 AM, Andre Merzky wrote:
> >Hi again,
> >
> >consider following use case for remote IO. Given a large
> >binary 2D field on a remote host, the client wans to access
> >a 2D sub portion of that field. Dependend on the remote
> >file layout, that requires usually more than one read
> >operation, since the standard read (offset, length) is
> >agnostic to the 2D layout.
> >
> >For more complex operations (subsampling, get a piece of a
> >jpg file), the number of remote operations grow very fast.
> >Latency then stringly discourages that type of remote IO.
> >
> >For that reason, I think that the remote file IO as
> >specified by SAGA's Strawman as is will only be usable for a
> >limited and trivial set of remote I/O use cases.
> >
> >There are three (basic) approaches:
> >
> > A) get the whole thing, and do ops locally
> > Pro: - one remote op,
> > - simple logic
> > - remote side doesn't need to know about file
> > structure
> > - easily implementable on application level
> > Con: - getting the header info of a 1GB data file comes
> > with, well, some overhead ;-)
> >
> > B) clustering of calls: do many reads, but send them as a
> > single request.
> > Pro: - transparent to application
> > - efficient
> > Con: - need to know about dependencies of reads
> > (a header read needed to determine size of
> > field), or included explicite 'flushes'
> > - need a protocol to support that
> > - the remote side needs to support that
> >
> > C) data specific remote ops: send a high level command,
> > and get exactly what you want.
> > Pro: - most efficient
> > Con: - need a protocol to support that
> > - the remote side needs to support that _specific_
> > command
> >
> >The last approach (C) is what I have best experiences with.
> >Also, that is what GridFTP as a common file access protocol
> >supports via ERET/ESTO operations.
> >
> >I want to propose to include a C-like extension to the File
> >API of the strawman, which basically maps well to GridFTP,
> >but should also map to other implementations of C.
> >
> >That extension would look like:
> >
> > void lsEModes (out array<string,1> emodes );
> > void eWrite (in string emode,
> > in string spec,
> > in string buffer
> > out long len_out );
> > void eRead (in string emode,
> > in string spec,
> > out string buffer,
> > out long len_out );
> >
> > - hooks for gridftp-like opaque ERET/ESTO features
> > - spec: string for pattern as in GridFTP's ESTO/ERET
> > - emode: string for ident. as in GridFTP's ESTO/ERET
> >
> >EMode: a specific remote I/O command supported
> >lsEModes: list the EModes available in this implementation
> >eRead/eWrite: read/write data according to the emode spec
> >
> >Example (in perl for brevity):
> >
> > my $file = SAGA::File new
> >("http://www.google.com/intl/en/images/logo.gif");
> > my @emodes = $file->lsEModes ();
> >
> > if ( grep (/^jpeg_block$/, @emodes) )
> > {
> > my ($buff, $len) = file.eRead ("jpeg_block", "22x4+7+8");
> > }
> >
> >I would discourage support for B, since I do not know any
> >protocoll supporting that approach efficiently, and also it
> >needs approximately the same infrastructure setup as C.
> >
> >As A is easily implementable on application level, or within
> >any SAGA implementation, there is no need for support on API
> >level -- however, A is insufficient for all but some trivial
> >cases.
> >
> >Comments welcome :-))
> >
> >Cheers, Andre.
> >
> >
> >--
> >+-----------------------------------------------------------------+
> >| Andre Merzky | phon: +31 - 20 - 598 - 7759 |
> >| Vrije Universiteit Amsterdam (VU) | fax : +31 - 20 - 598 - 7653 |
> >| Dept. of Computer Science | mail: merzky at cs.vu.nl |
> >| De Boelelaan 1083a | www: http://www.merzky.net |
> >| 1081 HV Amsterdam, Netherlands | |
> >+-----------------------------------------------------------------+
> >
--
+-----------------------------------------------------------------+
| Andre Merzky | phon: +31 - 20 - 598 - 7759 |
| Vrije Universiteit Amsterdam (VU) | fax : +31 - 20 - 598 - 7653 |
| Dept. of Computer Science | mail: merzky at cs.vu.nl |
| De Boelelaan 1083a | www: http://www.merzky.net |
| 1081 HV Amsterdam, Netherlands | |
+-----------------------------------------------------------------+
More information about the saga-rg
mailing list