On Mon, 2021-03-29 at 13:20 -0500, Benjamin Marzinski wrote: > > > > multipathd could figure out the system configuration from the (non) > > availability of certain properties, and use an appropriate fallback > > logic for either case. > > That seems like reasonable first step, although one that won't help > SID, > since it can't rely on getting the wwid from udev. Can you conceive of a different approach that would be better for SID? I'd like to hear about it. > This actually brings > up a different point I have. Is your main objection to adding more > config options that it is complicating the code, or confusing the > users? Both, with emphasis on the latter. I'm quite positive that we have too many options already. That doesn't mean I would generally oppose new options if they make sense. We should rather try to get rid of some new ones that nobody uses any more. > Because multipath wouldn't need to add any configuration options to > be > easily usable with SID (the current workaround, setting uid_attribute > to > "", is pretty non-obvious to users) if it could just check if sid was > running, and key off that. However this adds even more code > complexity > than simply checking a config option. I don't know how you would feel > about accepting a patch that does this, when SID is production ready. I could live well with this autodetection. I think it would be better than doing the same thing with a configuration options. Regards, Martin -- Dr. Martin Wilck <mwilck@xxxxxxxx>, Tel. +49 (0)911 74053 2107 SUSE Software Solutions Germany GmbH HRB 36809, AG Nürnberg GF: Felix Imendörffer -- dm-devel mailing list dm-devel@xxxxxxxxxx https://listman.redhat.com/mailman/listinfo/dm-devel
