2010/10/20 Daniel Veillard <veillard@xxxxxxxxxx>: > On Wed, Oct 20, 2010 at 02:37:53PM +0200, Matthias Bolte wrote: >> There is no point in trying to fill params beyond the first error, >> because when qemuDomainGetMemoryParameters returns -1 then the caller >> cannot detect which values in params are valid. >> --- > > ÂOkay ACK Thanks, pushed. >> There is a similar pattern in qemuDomainSetMemoryParameters that tries >> to apply all given params even if one already failed. From a user's >> POV it's probably better to apply all params without an error or to >> apply non at all when one fails. But this is harder to implemeneted >> and requires a rollback mechanism. > > ÂWell I don't think we can implement transactions there, so there will > always a case for uncertainties. Still since the client may want to tune > his guest domain, maybe some of the tuning will fail (and with the > current API, only retrying with one setting at a time can allow to > isolate the problem), but after the call we get closer to the expected > setup. It's still possible to see what was done with the Get API though. > ÂThe simplest way from a client perspective is probably to change > only one setting at a time, it's also likely to match the UI offered > to the end user. > > Daniel > Okay, that makes sense. Matthias -- libvir-list mailing list libvir-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/libvir-list