login
Header Space

 
 

Re: HAMMER update 06-Feb-2008

Previous thread: Re: HAMMER update 06-Feb-2008 by Matthew Dillon on Wednesday, February 6, 2008 - 9:40 pm. (1 message)

Next thread: Re: HAMMER update 06-Feb-2008 by Matthew Dillon on Thursday, February 7, 2008 - 3:31 pm. (1 message)
To: <kernel@...>
Date: Thursday, February 7, 2008 - 1:18 am

There were two issues.  First there were numerous buffer dependancies
    that required a synchronous write to 'open' the cluster before the rest
    of the buffers could be flushed.  It got messy really fast.

    The second issue was that I found it really hard to predict the key range
    for a new cluster to get a reasonable fill ratio before HAMMER would have

    Hmm.  I think the answer is no, historical access must be turned on
    (especially now with the FIFO mechanic).  There's no random bitmap free
    any more so 'nohistory' won't work anyway in the initial release.

    In otherwords, if your disk fills up the only way to clear space is to
    cycle the fifo.  That will ultimately be improved by a future backing
    store behind the fifo (the blockmap idea I talked about before), and/or
    out of band data references, allowing some immediate gratification for
    e.g. large files. 

    Personally speaking I think the blockmap idea may be the best way to go.
    It would be theoretically possible to immediately reuse a block that
    becomes completely free by unmapping it from the fifo (creating a 'hole'
    in the fifo) and freeing it for reuse.  One would thus not have to cycle
    the fifo to reach the block in order to reuse it.

					-Matt
					Matthew Dillon 
					&lt;dillon@backplane.com&gt;
To: <kernel@...>
Date: Thursday, February 7, 2008 - 3:38 pm

I might be wrong (and have misunderstood your reply) but I think he
asked whether it would be possible to disallow a user to access the
historical data (even if it is present on the disk). As I have not
studied the code I might be wrong, but should it not be quite easy to
create some switch which determines whether a @@0x&lt;timestamp&gt; will be
treated as a normal filename or the name or not?

-- 
Erik Wikström
To: <kernel@...>
Date: Thursday, February 7, 2008 - 4:07 pm

No. That would be, as you say, trivial :)

Aggelos
To: <kernel@...>
Date: Thursday, February 7, 2008 - 5:00 am

Now combine that with 64bit block idents and you have a "quasi infinte 
FIFO" :)

cheers
   simon

-- 
Serve - BSD     +++  RENT this banner advert  +++    ASCII Ribbon   /"\
Work - Mac      +++  space for low €€€ NOW!1  +++      Campaign     \ /
Party Enjoy Relax   |   http://dragonflybsd.org      Against  HTML   \
Dude 2c 2 the max   !   http://golden-apple.biz       Mail + News   / \
Previous thread: Re: HAMMER update 06-Feb-2008 by Matthew Dillon on Wednesday, February 6, 2008 - 9:40 pm. (1 message)

Next thread: Re: HAMMER update 06-Feb-2008 by Matthew Dillon on Thursday, February 7, 2008 - 3:31 pm. (1 message)
speck-geostationary