Repo Sync keeps on freezing..any ways to resume it!!?!

1,448 views
Skip to first unread message

Vince

unread,
Dec 28, 2011, 12:12:50 PM12/28/11
to Repo and Gerrit Discussion
Is there ANY WAY to resume the sync??? I literally tried it 8 times
and wasted almost 2 weeks just to download the source...and I keep on
restarting it back to zero!!! when I looked at the current directory
where it's supposed to be saved...there's nothing on it. I'm already
annoyed by this because this is really such a HUGE WASTE of time I've
been trying to be patient but all it gives me is headache because it
keeps on freezing! I never encountered this on downloading
torrents...EVEN downloading through the normal browser...what's
annoying me is that it keeps on going back to zero...and I have to
freakin start downloading from the very freakin start!

Shawn Pearce

unread,
Dec 28, 2011, 12:22:33 PM12/28/11
to Vince, Repo and Gerrit Discussion
On Wed, Dec 28, 2011 at 09:12, Vince <vince0v...@gmail.com> wrote:
> Is there ANY WAY to resume the sync???

repo sync can be resumed. Just cd into the directory and run `repo sync` again.

> I literally tried it 8 times
> and wasted almost 2 weeks just to download the source...and I keep on
> restarting it back to zero!!! when I looked at the current directory
> where it's supposed to be saved...there's nothing on it.

Next time try `ls -a`. There is a directory called ".repo" which
contains the information that has already been downloaded. Its hidden
because its version control information for the Git version control
system, it isn't user readable or modifiable. Once all data has been
downloaded, `repo sync` will then populate the checked out files that
are human readable.

> I'm already
> annoyed by this because this is really such a HUGE WASTE of time I've
> been trying to be patient but all it gives me is headache because it
> keeps on freezing!

Where does it freeze? The download is several GBs, and requires
several hundred HTTP requests to complete. Large portions should be
byte-range resumeable if the connection is disconnected.

> I never encountered this on downloading
> torrents...

I wonder if the torrent client you are using has an aggressive retry
policy that masks network problems? `repo sync` doesn't. It aborts and
assumes that retrying isn't recoverable without human intervention
(e.g. your network cable got unplugged because your cat tripped over
it). But as I said above, you can manually restart `repo sync`.

I think the exit status is correct, so you may be able to use a shell loop:

while true; do repo sync && break; sleep 10; done

Vince

unread,
Dec 28, 2011, 12:27:17 PM12/28/11
to Repo and Gerrit Discussion
How do I resume the sync?? I keep on going back to zero even after
cd'ing into the directory...

On Dec 29, 1:22 am, Shawn Pearce <s...@google.com> wrote:

Magnus Bäck

unread,
Dec 28, 2011, 12:31:44 PM12/28/11
to Repo and Gerrit Discussion
On Wednesday, December 28, 2011 at 18:27 CET,
Vince <vince0v...@gmail.com> wrote:

> How do I resume the sync?? I keep on going back to zero even after
> cd'ing into the directory...

That would indicate that you're not able to sync a single git. Can you
please post the output of a sync command?

--
Magnus Bäck Opinions are my own and do not necessarily
SW Configuration Manager represent the ones of my employer, etc.
Sony Ericsson

Vince

unread,
Dec 28, 2011, 12:34:36 PM12/28/11
to Repo and Gerrit Discussion
okay I saw the .repo folder..and it's a whopping 1.8GB in size...but
when I repo sync..it goes back to zero ??? and I already closed the
terminal window...

On Dec 29, 1:31 am, Magnus Bäck <magnus.b...@sonyericsson.com> wrote:
> On Wednesday, December 28, 2011 at 18:27 CET,

Magnus Bäck

unread,
Dec 28, 2011, 12:40:58 PM12/28/11
to Repo and Gerrit Discussion
On Wednesday, December 28, 2011 at 18:34 CET,
Vince <vince0v...@gmail.com> wrote:

> okay I saw the .repo folder..and it's a whopping 1.8GB in size...but
> when I repo sync..it goes back to zero ??? and I already closed the
> terminal window...

I'm sure you can find a way to open a new terminal window and retry
the command.

If your .repo directory is 1.8 GB in size your sync is certainly
working. Let's see some proof that it's actually restarting every
time. You do realize that there are a couple of hundred gits that
need to be fetched, and that each new git will start at zero?

Reply all
Reply to author
Forward
0 new messages