From: David Turner <David.Turner@twosigma.com>
To: 'Jeff King' <peff@peff.net>
Cc: "git@vger.kernel.org" <git@vger.kernel.org>,
"christian.couder@gmail.com" <christian.couder@gmail.com>,
"mfick@codeaurora.org" <mfick@codeaurora.org>,
"jacob.keller@gmail.com" <jacob.keller@gmail.com>
Subject: RE: [PATCH] repack: respect gc.pid lock
Date: Thu, 20 Apr 2017 20:10:24 +0000 [thread overview]
Message-ID: <7e31f4ed5c0f4c31b2870fb58cf7110e@exmbdft7.ad.twosigma.com> (raw)
In-Reply-To: <20170418175011.qx64luolrvqwwtpa@sigill.intra.peff.net>
> -----Original Message-----
> From: Jeff King [mailto:peff@peff.net]
> Sent: Tuesday, April 18, 2017 1:50 PM
> To: David Turner <David.Turner@twosigma.com>
> Cc: git@vger.kernel.org; christian.couder@gmail.com; mfick@codeaurora.org;
> jacob.keller@gmail.com
> Subject: Re: [PATCH] repack: respect gc.pid lock
>
> On Tue, Apr 18, 2017 at 05:43:29PM +0000, David Turner wrote:
>
> > > A lock can catch the racy cases where both run at the same time. But
> > > I think that
> > > even:
> > >
> > > git -c repack.writeBitmaps=true repack -Ad
> > > [...wait...]
> > > git gc
> > >
> > > is questionable, because that gc will erase your bitmaps. How does
> > > git-gc know that it's doing a bad thing by repacking without
> > > bitmaps, and that you didn't simply change your configuration or want to get
> rid of them?
> >
> > Sorry, the gc in Gitlab does keep bitmaps. The one I quoted in a
> > previous message doesn't, because the person typing the command was
> > just doing some manual testing and I guess didn't realize that
> > bitmaps were important. Or perhaps he knew that repack.writeBitmaps was
> already set in the config.
>
> Sure, but I guess I'd just wonder what _else_ is different between the commands
> (and if nothing, why are both running).
Presumably, repack is faster, and they're not intended to run concurrently (but
there's a Gitlab bug causing them to do so). But you'll have to ask the Gitlab
folks for more details.
> > So given that the lock will catch the races, might it be a good idea
> > (if Implemented to avoid locking on repack -d)?
>
> I'm mildly negative just because it increases complexity, and I don't think it's
> actually buying very much. It's not clear to me which invocations of repack
> would want to lock and which ones wouldn't.
>
> Is "-a" or "-A" the key factor? Are there current callers who prefer the current
> behavior of "possibly duplicate some work, but never report failure" versus "do
> not duplicate work, but sometimes fail due to lock contention"?
One problem with failing is that it can leave a temp pack behind.
I think the correct fix is to change the default code.packedGitLimit on 64-bit
machines to 32 terabytes (2**45 bytes). That's because on modern Intel
processors, there are 48 bits of address space actually available, but the kernel
is going to probably reserve a few bits. My machine claims to have 2**46 bytes
of virtual address space available. It's also several times bigger than any
repo that I know of or can easily imagine.
Does that seem reasonable to you?
next prev parent reply other threads:[~2017-04-20 20:11 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-04-13 20:27 [PATCH] repack: respect gc.pid lock David Turner
2017-04-14 0:33 ` Jacob Keller
2017-04-14 19:33 ` Jeff King
2017-04-17 23:29 ` David Turner
2017-04-18 3:41 ` Jeff King
2017-04-18 17:08 ` David Turner
2017-04-18 17:16 ` Jeff King
2017-04-18 17:16 ` David Turner
2017-04-18 17:19 ` Jeff King
2017-04-18 17:43 ` David Turner
2017-04-18 17:50 ` Jeff King
2017-04-20 20:10 ` David Turner [this message]
2017-04-20 20:14 ` Jeff King
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
List information: http://vger.kernel.org/majordomo-info.html
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=7e31f4ed5c0f4c31b2870fb58cf7110e@exmbdft7.ad.twosigma.com \
--to=david.turner@twosigma.com \
--cc=christian.couder@gmail.com \
--cc=git@vger.kernel.org \
--cc=jacob.keller@gmail.com \
--cc=mfick@codeaurora.org \
--cc=peff@peff.net \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
Code repositories for project(s) associated with this public inbox
https://80x24.org/mirrors/git.git
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).