On Dienstag, 1. Mai 2007, Norberto Bensa wrote: > Hemmann, Volker Armin wrote: > > no, I trust the kernel devs how said, that a swapfile in 2.6 should not > > be slower. > > "should not" is not the same as "is not" > > I better trust my own experiences...
and I mine - which showed me in the past, that both suck the same. > > > > /dev/swap-device none swap sw 0 0 > > > > > > (Note mount-point is none.) > > > > you 'activate it', which is pretty similar to mounting. > > Yup. And a swapfile needs mounting and activating. > and? What is the problem? The partition has to be mounted anyway, and activating a swapfile or swap partition takes the same amount of time. swapfiles even have the big advantage, that you can remove or resize them easily. Need more space? Remove it! Need more swap? Create another one. I do have a swap partition ATM, because I abuse it for bios updates once in a while (converting it to fat and use it to store the files, so I can boot from a freedos cd), and a swapfile for emergencies. -- [EMAIL PROTECTED] mailing list
