| 1 |
On Wed, Aug 24, 2011 at 11:48:05AM +0300, Eray Aslan wrote: |
| 2 |
> On Wed, Aug 24, 2011 at 07:30:11AM +0000, Robin H. Johnson wrote: |
| 3 |
> > Bandwidth: Along the same lines, rsync will always be able to use less |
| 4 |
> > bandwidth than Git, because none of the intermediate commits need to be |
| 5 |
> > transfered. This will be esp. evident as a user tree gets older (the |
| 6 |
> > amount of mtime/checksum metadata scales linearly with the size of the |
| 7 |
> > tree, not the age of the tree. |
| 8 |
> Yes, rsync protocol scales with the project size. |
| 9 |
> > The actual file content transfered scales linearly with the age of the tree). |
| 10 |
> (assuming you are talking about git here) |
| 11 |
> Uhm no? git protocol scales with chageset size. You don't retransfer |
| 12 |
> already transferred content. So you will be tranferring only content |
| 13 |
> from the last git pull. |
| 14 |
No, I'm talking about rsync, not Git. |
| 15 |
|
| 16 |
Rsync bandwidth usage comprises two parts: |
| 17 |
1. Bandwidth used by the up-to-date-checks. |
| 18 |
Scales with the size of the tree. |
| 19 |
2. Bandwidth used by sending newest versions of files. |
| 20 |
Scales with the number of changed files. |
| 21 |
|
| 22 |
Let's call this B(checked) + B(changed) |
| 23 |
|
| 24 |
However, most importantly, it's that rsync only sends the NEWEST version |
| 25 |
of the files, not the complete history. |
| 26 |
|
| 27 |
Git bandwidth usage comprises one thing, the set of commits since the |
| 28 |
last time you did a fetch. |
| 29 |
|
| 30 |
Let's call this B(commits) |
| 31 |
|
| 32 |
If you run rsync very frequently, B(changed) tends towards zero, |
| 33 |
while B(checked) becomes constant. |
| 34 |
|
| 35 |
If you run git very frequently, B(commits) tends towards zero. |
| 36 |
|
| 37 |
Now, the important thing happens when you run them seldom. |
| 38 |
The sum of B(changed)+B(checked) will become smaller than B(commits) |
| 39 |
when there are many commits that have been superseded by newer commits. |
| 40 |
|
| 41 |
Yes, there are cases where EITHER side can win, but for the rate of |
| 42 |
change of the tree and how often rsync should be used (not more than |
| 43 |
every 12 hours), I believe that the rsync bandwidth usage will be less |
| 44 |
than that of Git. |
| 45 |
|
| 46 |
-- |
| 47 |
Robin Hugh Johnson |
| 48 |
Gentoo Linux: Developer, Trustee & Infrastructure Lead |
| 49 |
E-Mail : robbat2@g.o |
| 50 |
GnuPG FP : 11AC BA4F 4778 E3F6 E4ED F38E B27B 944E 3488 4E85 |
Archives