git@vger.kernel.org mailing list mirror (one of many)
 help / color / mirror / code / Atom feed
From: 16657101987@163.com
To: unlisted-recipients:; (no To-header on input)
Cc: git@vger.kernel.org, gitster@pobox.com, worldhello.net@gmail.com,
	Sun Chao <sunchao9@huawei.com>
Subject: [PATCH v1 0/1] pack-refs: pack expired loose refs to packed_refs
Date: Mon, 22 Jul 2019 02:17:38 +0800	[thread overview]
Message-ID: <20190721181739.81110-1-16657101987@163.com> (raw)

From: Sun Chao <sunchao9@huawei.com>

When a packed ref is deleted, the whole packed-refs file is
rewrite and omit the ref that no longer exists. However if
another gc command is running and call `pack-refs --all`
simultaneously, there is a change that a ref just updated
will lost the newly commits.

There are two valid methods to avoid losting commit of ref:
  - force `update-ref -d` to update the snapshot before
    rewrite packed-refs.
  - do not pack a recently updated ref, where *recently*
    could be set by *pack.looserefsexpire* option.

I prefer **do not pack a recently updated ref**, here is the
reasons:

  1. It could avoid losting the newly commit of a ref which I
     described upon.

  2. Sometime, the git server will do `pack-refs --all` and
     `update-ref` the same time, and the two commands have
     chances to trying lock the same ref such as master, if
     this happends one command will fail with such error:

     **cannot lock ref 'refs/heads/master'**

     This could happen if a ref is updated frequently, and
     avoid pack the ref which is update recently could avoid
     this error most of the time.

Sun Chao (1):
  pack-refs: pack expired loose refs to packed_refs

 builtin/pack-refs.c       | 13 ++++++++++++-
 refs.c                    |  4 ++--
 refs.h                    |  2 +-
 refs/files-backend.c      | 18 +++++++++++++++++-
 refs/packed-backend.c     |  2 +-
 refs/refs-internal.h      |  2 +-
 t/helper/test-ref-store.c |  2 +-
 7 files changed, 35 insertions(+), 8 deletions(-)

-- 
2.22.0.214.g8dca754b1e



             reply	other threads:[~2019-07-21 18:18 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-07-21 18:17 16657101987 [this message]
2019-07-21 18:17 ` [PATCH v1 1/1] pack-refs: pack expired loose refs to packed_refs 16657101987
2019-07-30  6:36   ` Jeff King
2019-07-31 18:35     ` [PATCH v2 0/1] pack-refs: always refreshing after take the lock file 16657101987
2019-07-31 18:35       ` [PATCH v2 1/1] " 16657101987
2019-08-16 20:49       ` [PATCH v2 0/1] " Jeff King
2019-08-19 17:36         ` Junio C Hamano
2019-08-20 15:14           ` 16657101987

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=20190721181739.81110-1-16657101987@163.com \
    --to=16657101987@163.com \
    --cc=git@vger.kernel.org \
    --cc=gitster@pobox.com \
    --cc=sunchao9@huawei.com \
    --cc=worldhello.net@gmail.com \
    /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).