git/list[1] front-page[2] threads[3] people[4] search[5] about
 

Re: Resolving deltas dominates clone time

From
MFMartin Fick <mfick@codeaurora.org>
Date
Apr 30, 2019, 22:08 UTC
Message-ID
<7252129.tRx6dl7m8I@mfick-lnx>
In-Reply-To
<20190430180231.GC16729@sigill.intra.peff.net>
On Tuesday, April 30, 2019 2:02:32 PM MDT Jeff King wrote:
Show 14 quoted lines
> On Tue, Apr 23, 2019 at 02:09:31PM -0600, Martin Fick wrote:
> > I think that if there were no default limit during a clone it could have
> > disastrous effects on people using the repo tool from the android project,
> > or any other "submodule like" tool that might clone many projects in
> > parallel. With the repo tool, people often use a large -j number such as
> > 24 which means they end up cloning around 24 projects at a time, and they
> > may do this for around 1000 projects. If git clone suddenly started as
> > many threads as there are CPUs for each clone, this would likely paralyze
> > the machine.
> 
> IMHO this is already a problem, because none of those 24 gits knows
> about the others. So they're already using 24*3 cores, though of course
> at any given moment some of those 24 may be bottle-necked on the
> network.

I think this is very different. In the first case this is a linear constraint that someone can use to intelligently adjust the number of clones they are doing at the same time. The second case cannot be accounted for in any intelligent way by the person running the clones, it is almost unconstrained.

...
Show 12 quoted lines
> > I do suspect it would be nice to have a switch though that repo could use
> > to adjust this intelligently, is there some way to adjust threads from a
> > clone, I don't see one? I tried using 'GIT_FORCE_THREADS=28 git clone
> > ...' and it didn't seem to make a difference?
> 
> I think I led you astray earlier by mentioning GIT_FORCE_THREADS. It's
> actually just a boolean for "use threads even if we're only
> single-threaded". What you actually want is probably:
> 
>   git clone -c pack.threads=28 ...
> 
> (though I didn't test it to be sure).
Thanks, I will test this!
-Martin
-- 
The Qualcomm Innovation Center, Inc. is a member of Code 
Aurora Forum, hosted by The Linux Foundation
Previous: Jeff KingNext: Jeff King
Message 23 of 26 in “Resolving deltas dominates clone time”
  1. Martin FickApr 19, 2019
  2. Jeff KingApr 20, 2019
  3. Ævar Arnfjörð BjarmasonApr 20, 2019
  4. Jeff KingApr 22, 2019
  5. Ævar Arnfjörð BjarmasonApr 22, 2019
  6. Jeff KingApr 22, 2019
  7. Ævar Arnfjörð BjarmasonApr 23, 2019
  8. Martin FickApr 22, 2019
  9. Jeff KingApr 22, 2019
  10. Jeff KingApr 22, 2019
  11. p5302: create the repo in each index-pack testJeff King, Apr 22, 2019
  12. Junio C HamanoApr 23, 2019
  13. Jeff KingApr 23, 2019
  14. Junio C HamanoApr 23, 2019
  15. Jeff KingApr 23, 2019
  16. Junio C HamanoApr 23, 2019
  17. Martin FickApr 22, 2019
  18. Jeff KingApr 23, 2019
  19. Jeff KingApr 23, 2019
  20. Duy NguyenApr 23, 2019
  21. Martin FickApr 23, 2019
  22. Jeff KingApr 30, 2019
  23. Martin FickApr 30, 2019
  24. Jeff KingApr 30, 2019
  25. Ævar Arnfjörð BjarmasonApr 30, 2019
  26. Jeff KingApr 30, 2019

Read the whole thread, see it on lore, or plain text.

$ cat FOOTERMessages come from the public archive at lore.kernel.org/git, fetched every hour. The front page is chosen and written each morning by an AI editor and can be wrong; the threads themselves are the record. About and API. For agents: an MCP server at https://gitlist.dev/mcp, and any thread, story or person page as Markdown by adding .md to its URL (or sending Accept: text/markdown). Details in /llms.txt.