| 1 |
On Fri, Jun 1, 2012 at 4:49 PM, Michael Weber <xmw@g.o> wrote: |
| 2 |
> |
| 3 |
> Cloning the repo [1] takes 200seconds on 8cores (it's 2GB of data and |
| 4 |
> 22 minutes of 3.4GHz cpus). |
| 5 |
> |
| 6 |
|
| 7 |
As others have pointed out, probably the best way to bootstrap this is |
| 8 |
to offer tarballs of a shallow repository and a full repository. |
| 9 |
Perhaps we'd offer the latter as a torrent. The shallow repository |
| 10 |
should be light on the CPU too. |
| 11 |
|
| 12 |
This would be a lot easier on the server than having everybody and |
| 13 |
their uncle doing a full 2GB clone. |
| 14 |
|
| 15 |
Devs could then do a pull to get the latest and greatest, and that |
| 16 |
would only transfer the delta. |
| 17 |
|
| 18 |
I imagine our mirror network can handle the bandwidth compared to what |
| 19 |
we're already doing with distfiles. Worst case we could take a |
| 20 |
one-time hit and use S3 or whatever to do the distribution (they even |
| 21 |
support bittorrent to cut down on the bill). |
| 22 |
|
| 23 |
Rich |
Archives