Hi Heiner, On Monday 19 November 2007 14:23, Jens-Heiner Rechtien wrote:
> > I tried to svnsync the repository over http. The results were quite > > disappointing, after about one hour runtime, I had only fetched 17 (yes, > > seventeen) changesets. As far as I can see the repository contains over > > 500000 changesets. > > > > During that time, I had always plenty of CPU% available and the network > > (DSL-2000) also had a lot of capacity left. > > > > After one hour I decided to stopped it. My local repo with the 17 > > changesets now has a 'du -k' size of 69356. > > > > Do you have an explanation for this bad behaviour? > > The first 20 or so changesets contains extraordinarily much stuff, as it > represents they first complete import in CVS in 2000. That said, the > repository is still huge. It's probably bigger than necessary, but a > real repository will grow anyway. > > 69 MBytes in one hour over DSL2000 is not good enough (no matter which > SCM we choose). I'll check if the apache2 on the server might be the > bottleneck. So far, my svn synchronization which I started on Friday is at about revision 117000 (it failed over the weekend, and I restarted it today in the morning); so does not look too optimistic either :-( [It's over a faster line than DSL]. Regards, Jan --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
