tech-repository archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: "hg clone https://anonhg.netbsd.org/src/" still aborts, but...



At Fri, 12 Jun 2020 13:19:29 -0400, John Franklin <franklin%elfie.org@localhost> wrote:
Subject: Re: "hg clone https://anonhg.netbsd.org/src/"; still aborts, but...
>
> How much RAM do you have in that machine?  Could some of that time
> waffling be due to the system thrashing to swap?  My experience with
> doing similar tests is hg required >3GB of RAM+swap to complete, and
> Git required about one-quarter the memory.

It has WAY more than enough RAM.  8GB  (and with 8x3.16GHz cores)

I've _never_ seen this machine push anything more than truly idle unused
pages to swap (i.e. it has never thrashed in any way), and that's with
two parallel build.sh runs (running a fully static-linked userland and
toolchain), and also with a couple of large emacs processes, often one
that is >1GB, and I watch it like a hawk with "systat vm" and/or "top"
at times when it is most busy.

The time to do the checkout is clearly limited by the NFS server
(especially since it's on an entirely separate physical machine with a
1GB ethernet link), but that's in theory at least exactly the same
amount of I/O and data that Git has to move to do the same checkout.  HG
is just way slower, at least when doing the primary bundle unpacking
during the clone operation and again during the initial "hg checkout".

The "hg pull -u" to update from the state of the bundle was fast enough
(this was started directly after the initial "hg unbundle && hg checkout
trunk").  By comparison a "git pull" takes a similar amount of time.

$ time hg pull -u
pulling from https://anonhg.netbsd.org/src
searching for changes
adding changesets
adding manifests
adding file changes
added 1288 changesets with 8523 changes to 5257 files (+1 heads)
adding remote bookmark @
new changesets 26c8f37631b6:57842fb6f75e
931876 local changesets published
5222 files updated, 0 files merged, 447 files removed, 0 files unresolved
updated to "57842fb6f75e: pmf-powerdown == false  -->  pmf-no-powerdown == true"
1 other heads for branch "trunk"
  556.58s real    42.92s user    74.93s system


--
					Greg A. Woods <gwoods%acm.org@localhost>

Kelowna, BC     +1 250 762-7675           RoboHack <woods%robohack.ca@localhost>
Planix, Inc. <woods%planix.com@localhost>     Avoncote Farms <woods%avoncote.ca@localhost>

Attachment: pgp2FYiaz0wD6.pgp
Description: OpenPGP Digital Signature



Home | Main Index | Thread Index | Old Index