Re: nfs - date & time (settingdelay?) -
David C Rankin <[email protected]>
| Newsgroups | gmane.linux.suse.general |
|---|---|
| Message-ID | <[email protected]> |
On 8/27/26 1:03 PM, Simon Becherer via openSUSE Users wrote: <snip> > cat /sys/kernel/debug/nfsd/delegated_timestamps > echo 0 > /sys/kernel/debug/nfsd/delegated_timestamps > > after this the time of the files will be set correct and not switch after > a short time or set completely wrong. > > the second problem is also analysed with help of ai, if the copy over > nfs will > be done with nfsv3 (and not with nfsv4) there is no delay, so with the > global switch > echo 0 > /proc/sys/fs/leases-enable > there is also no delay with nfsv4 but maby switch off globally is not so > good. > > i reported 2 bugs: Bugzilla – Bug 1277253 Submitted > > Bugzilla – Bug 1277255 Submitted > Great job tracking down the issue and opening the bug reports. Seems AI is good for something. You feed it enough data, scrape every corner of the internet, disregard copyright and then buy all the used, historical and cultural books in every language you can find, destructively rip the bindings off, and feed them into a scanner to get eveery last scrap of information, destroying culture and heritage along the way without preserving the text, and balloon energy consumption by TWs (TeraWatts, not Tumbleweeds) of compute emitting gigatons of CO2 causing glacier collapse in the Himalayas -- and AI makes a pretty good search engine for solving computer problems.... It will be interesting to see what comes out of the bug reports and why the defaults are what you found. That's frustrating when it, e.g. "works fine with nfsv3" but not with nfsv4. That also reminds me I have an old tower in the bone-pile that still has SUSE 9.0 Pro as the OS. I don't think I've booted it in 20? years. But unless the caps have gone on the MB, I bet it starts right up, on a system with a whopping 256MB of RAM :) -- David C. Rankin, J.D.,P.E.