Date: Tue, 11 Aug 2015 12:38:45 -0300 From: Christopher Forgeron <csforgeron@gmail.com> To: Karl Denninger <karl@denninger.net> Cc: FreeBSD Filesystems <freebsd-fs@freebsd.org> Subject: Re: FreeBSD 10.1 Memory Exhaustion Message-ID: <CAB2_NwCgGfoXfK%2BUAxC6CGhw9rGyFy5Y%2BS5P_ALM8c1ttK%2BdPQ@mail.gmail.com> In-Reply-To: <55A6C4BD.5090500@denninger.net> References: <CAB2_NwCngPqFH4q-YZk00RO_aVF9JraeSsVX3xS0z5EV3YGa1Q@mail.gmail.com> <55A3A800.5060904@denninger.net> <55A4D5B7.2030603@freebsd.org> <55A4E5AB.8060909@netlabs.org> <CACfj5vJvAz9StvjTrA1TzfS%2BMhi_qSrOc_qBNHr8qXbiAj81xw@mail.gmail.com> <1436989410.1427298.324703241.421E814B@webmail.messagingengine.com> <55A6C0A4.1030300@digiware.nl> <55A6C4BD.5090500@denninger.net>
next in thread | previous in thread | raw e-mail | index | archive | help
Hello All, Any reports on success / issues with the patch? I'm reporting a good increase in uptime and stability. I have two production machines with heavy load running, and I've made it almost a month without a crash on one machine. The other had a swap-related crash that wasn't really a fault/issue of the patch, but even with that one crash, it's been more stable than a non-patched system. I'm curious to everyone else's findings. On Wed, Jul 15, 2015 at 5:38 PM, Karl Denninger <karl@denninger.net> wrote: > > On 7/15/2015 15:20, Willem Jan Withagen wrote: > > On 15/07/2015 21:43, Mark Felder wrote: > >> > >> On Tue, Jul 14, 2015, at 10:10, Sean Chittenden wrote: > >>> I think the reason this is not seen more often is because people > >>> frequently > >>> throw limits on the arc in /boot/loader.conf: > >>> > >>> vfs.zfs.arc_min="18G" > >>> vfs.zfs.arc_max="149G" > >>> > >>> ZFS ARC *should* not require those settings, but does currently for > mixed > >>> workloads (i.e. databases) in order to be "stable". By setting fixed > >>> sizes > >>> on the ARC, UMA and ARC are much more cooperative in that they have > their > >>> own memory regions to manage so this behavior is not seen as often. > >>> > >>> To be clear, however, it should not be necessary to set parameters like > >>> these in /boot/loader.conf in order to obtain consistent operational > >>> behavior. I'd be curious to know if someone running 10.2 BETA without > >>> patches is able to trigger this behavior or not. There was work done > >>> that > >>> reported helped with this between 10.1 and now. To what extent it > >>> helped, > >>> however, I don't have any advice yet. > >>> > >> I was about to email "I have 12TB at home and 4GB of RAM with a very > >> erratic workload and never run into any issues" and then I looked at > >> /boot/loader.conf and saw vfs.zfs.arc_max="2G" > >> > >> Now I'm too scared to turn it off... :-) > > Same here. > > Just a leftover of all the advise to limit arc in the past. > > Just bit the bullit: installed BETA1, killed the settings and rebooted. > > We'll see what comes of it. > > > > --WjW > If you get bit I have refactored the patch for 10.2-BETA1 to get rid of > the fudges and i386 linker problem and uploaded it, so if you run into > trouble try applying that and see if that fixes it. > > https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=187594 > > -- > Karl Denninger > karl@denninger.net <mailto:karl@denninger.net> > /The Market Ticker/ > /[S/MIME encrypted email preferred]/ >
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CAB2_NwCgGfoXfK%2BUAxC6CGhw9rGyFy5Y%2BS5P_ALM8c1ttK%2BdPQ>