1649387740 M * Bertl_oO off to bed now ... have a good one everyone! 1649387741 N * Bertl_oO Bertl_zZ 1649391688 J * Ghislain ~ghislain@adsl2.aqueos.com 1649411470 N * Bertl_zZ Bertl 1649411484 M * Bertl morning folks! 1649430607 M * Bertl off for now ... bbl 1649430608 N * Bertl Bertl_oO 1649431683 Q * _are_ Read error: No route to host 1649431787 J * _are_ ~quassel@jabber.lihas.de 1649441840 M * Hurga Hi... I just noticed a slightly strage behavior. I have a long-running machine (uptime 443 days) with kernel 4.9.217-vs2.3.9.12 and several vserver guests. ps -ef shows some STIME weirdness. 1649441849 M * Hurga In the guests, I did 'date' followed by 'sleep 200 &' and then ps -ef | grep sleep in quick succession. Results are in https://www.toptal.com/developers/hastebin/omuvezudup.apache 1649441863 M * Hurga I can't make heads or tails of the differences to the real time... 1649441914 M * Hurga The guest uptimes are from vserver-stat 1649442076 M * Bertl_oO well, the guest share a clock with the host, but there is some offset stored per guest to make different 'up-times' possible 1649442181 M * Bertl_oO probably the first guest (ubuntu) was started shortly after the host booted up 1649442208 M * Bertl_oO so the delta (offset) is almost 0 1649442208 M * Hurga Inside the guest, the STIME should be the time of process creation, which diesn't have much to do with uptime 1649442234 M * Bertl_oO yeah, but the question is how this value is stored 1649442254 M * Bertl_oO it could easily be stored as the time since the system was booted ;) 1649442327 M * Bertl_oO the first step would be to check whether this behaviour is also present in the recent patches (which is rather likely) 1649442376 M * Hurga Will do, I need to set up a new machine anyway 1649442384 M * Bertl_oO then we need some replicable test case (which is probably quite easy with a few commands) 1649442448 M * Bertl_oO it would also be good to know the kernel and guest configuration (especially regarding the uptime virtualization) 1649442484 M * Bertl_oO and then we can probably figure out what needs to be adjusted to make this work as intended ;) 1649442584 M * Hurga I seem to remember there's a configuration option for virtalized uptimes, which I think means that inside the guest you see the uptime as the time the guest was running, instead of the uptime of the root server. That probably would make a difference too (I don't have it) 1649442622 M * Hurga I will set up a new machine and get back to you, ok? (any advice on current patch versions to use?) 1649442634 M * Bertl_oO yeah, so different options here might give different results 1649442697 M * Bertl_oO give the latest patch a try, let me know if it doesn't apply to the latest mainline kernel from that branch 1649442719 M * Bertl_oO in which case I'll update the patch 1649442725 M * Hurga ok, will do. 1649442747 M * Hurga Thanks for the quick help :) 1649442750 M * Bertl_oO and yes, please get back to me with the results! 1649442755 M * Bertl_oO you're welcome! 1649454136 Q * Ghislain Ping timeout: 480 seconds