On Saturday, March 10, 2018 at 1:21:25 AM UTC+1, [ 799 ] wrote: > Hello, > > > > Am 10.03.2018 1:10 vorm. schrieb "'MirrorWay' via qubes-users" > <qubes...@googlegroups.com>: > > You can reduce the start time to almost zero by using an already-running, > named DIspVM, see marmarek's post in > https://github.com/QubesOS/qubes-issues/issues/2801. > > > > > That sounds very interesting. > I have looked at the link, but didn't figure out what to do, to get faster > DispVM boot up times. > What do I need to do? > > > > > > > > You can set a cron job that ensures they shutdown at least once per day. > > > > Why? The DispVM should be shutdown after I close the window. > > > [799]
nice CPU/virt_mode/memory benchmarks! That was a really interesting read. btw I think what Mirrorway meant is if it automatically shutting down are to make up for down-time, for example if you don't use dispVM's for a full day or longer, then it'll shutdown on it's own and start again. Thereby, I believe, you prevent any potential internet based attacks by reloading fresh and clean system-files from the template. It seems like a pretty cool idea, it automates everything even if not using the computer for a period of time. Maybe make it more frequent, say once every 3 or 6 hours? btw I found this too just now https://github.com/QubesOS/qubes-issues/issues/2253 It seems even if the dispVM is shutdown, it can be made much faster too with a savefile. But if I understood it right, as Marek write in the first post they lack the manpower to get a savefile working for Qubes 4. But Qubes 3.2. has one, as it can be seen in Marek's time comparison in his second post. but whoa, in Qubes 3.2. a savefile makes a difference from 25,5 seconds to 4 seconds, on this particular hardware. Considering this is a completely shutdown dispVM, that is some pretty impressive speed differences. I wonder what would happen on Qubes 4 with PVH mode, savefile enabled, powerful CPU with a really fast NVMe SSD and RAM, all cores assigned but one which is kept in dom0, just how fast would it be then? In theory it should be less than 4 seconds at least if Marek's number on Qubes 3.2. can be seen as a maximum here for hardware, which is hardware that right now isn't the fastest available. What speeds would be possible here? Once a savefile is made for Qubes 4, this could probably be possible. -- You received this message because you are subscribed to the Google Groups "qubes-users" group. To unsubscribe from this group and stop receiving emails from it, send an email to qubes-users+unsubscr...@googlegroups.com. To post to this group, send email to qubes-users@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/qubes-users/32d92eb2-f5a2-4c62-a931-713fd227d78c%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.