Bug#876035: linux: i386: i/o with PAE kernels and > 8GB ram
Control: severity -1 minor Control: tag -1 upstream On Sun, 2017-09-17 at 10:16 -0700, Vagrant Cascadian wrote: [...] > This can be seen on the reproducible builds i386 nodes, after upgrading > to stretch late june/early july 2017, there was a huge spike in iowait: > > > https://jenkins.debian.net/munin/debian.net/profitbricks-build2-i386.debian.net/cpu.html > > https://jenkins.debian.net/munin/debian.net/profitbricks-build12-i386.debian.net/cpu.html > > > Switching to an amd64 kernel for hardware that supports it is an option > in many cases, though at least for reproducible builds purposes, doesn't > test reproducibility issues case by 32-bit vs. 64-bit kernel variations. This is just not a sensible configuration today. I'm reducing the severity accordingly. Ben. -- Ben Hutchings friends: People who know you well, but like you anyway. signature.asc Description: This is a digitally signed message part
Processed: Re: Bug#876035: linux: i386: i/o with PAE kernels and > 8GB ram
Processing control commands: > severity -1 minor Bug #876035 [src:linux] linux: i386: i/o with PAE kernels and > 8GB ram Severity set to 'minor' from 'important' > tag -1 upstream Bug #876035 [src:linux] linux: i386: i/o with PAE kernels and > 8GB ram Ignoring request to alter tags of bug #876035 to the same tags previously set -- 876035: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=876035 Debian Bug Tracking System Contact ow...@bugs.debian.org with problems
Bug#876035: linux: i386: i/o with PAE kernels and > 8GB ram
Source: linux Severity: important Version: 4.2 Tags: upstream Control: affects -1 jenkins.debian.org Control: block 875990 by -1 It seems that since linux 4.2, PAE systems suffer from severe i/o write performance issues on systems with more than 8GB of ram. The bug was probably there all along, but recent changes in the kernel made the impacts more obvious... https://bugzilla.kernel.org/show_bug.cgi?id=196157 https://bugs.launchpad.net/ubuntu/+source/linux-hwe/+bug/1698118 A detailed description of the issue: http://flaterco.com/kb/PAE_slowdown.html This can be seen on the reproducible builds i386 nodes, after upgrading to stretch late june/early july 2017, there was a huge spike in iowait: https://jenkins.debian.net/munin/debian.net/profitbricks-build2-i386.debian.net/cpu.html https://jenkins.debian.net/munin/debian.net/profitbricks-build12-i386.debian.net/cpu.html Switching to an amd64 kernel for hardware that supports it is an option in many cases, though at least for reproducible builds purposes, doesn't test reproducibility issues case by 32-bit vs. 64-bit kernel variations. live well, vagrant signature.asc Description: PGP signature