MPU Meeting Tuesday 16th December


The schema was updated to support all the latest features. Stephen tweaked the config for many standalone services so that the LCFG_SYSTEMD_UNIT_WANTEDBY macro was used. This results in files being created in the relevant .wants directory so that the services are started. Stephen also changed some of the systemd macros so that all resources are set using mutations, this makes the macros safe for multiple calls for the same unit.

Stephen will add a short note to the Systemd Cookbook explaining how to detect the init system using sysinfo.

SL7 LCFG port

Chris is continuing to work on the SL7 port of the sleep component and, in particular, making it work with systemd. He is also taking the chance to convert the templates over to using Perl TT. It looks like the latest versions of the requisite Perl modules now all build correctly on SL7.

Stephen applied a patch from Kenny for the Service function in LCFG::Component (see #810).

The cleanup of the mock and yum configurations is now completed and everything seems to be working a lot better. Stephen also took the chance to improve the Fedora and Centos templates which are used by the openafs build host, this brings in support for building packages on Fedora 21.

The network component docs now include a note stating the lack of support for SL7. The lvm component has been dropped from the SL7 installer.

The latest version of the SL7 installer has been built and we have dropped the hacky version Toby was using to test kdcregister.

More of the sssd/openldap changes have moved from the inf level to the relevant lcfg and dice headers.

Stephen looked into the problem with logrotate needing the "su" directive on SL7 because the log directory is group writable and the group is not root (see #812). He has added a new "features" resource to the sysinfo component which can take a list of flags which describe general features of the system. The sysinfo module has a new has_feature method to make it easy to test if a feature are enabled. For convenience, the shell version of ngeneric also has a new HasSysFeature function. The advantage here is that we do not need to test for OS version or for package version. This is particularly useful, given that Redhat often backport features without updating the major version of a package. Stephen also took the chance to introduce a new lcfg-sysinfo command-line query tool which can be called from any shell script. The shell version of ngeneric has been converted to using the new query tool which massively simplifies the code.

We will organise a planning meeting for early in the new year for all units so we can summarise current progress and what still needs to be done. We will aim for a "stable" release to external users in early January and the first DICE desktops for users at the start of March.

Miscellaneous Development

This package list has got rather large again. We need to shuffle lots of the packages into the core release-managed list.

ngeneric plugins
Stephen has implemented a new method environment initialisation system for Perl-based LCFG components (see EnvInit on the LCFG wiki. This supports plugins, there are currently two available - one for setting environment variables and one for getting Kerberos credentials along with AFS tokens. The rkhunter component is the first to use this new feature which should mean that it will always start automatically at boot-time rather than requiring manual intervention.

cron and AFS
Whilst working on SL7 Stephen discovered a bug in the cronie package which means that cron does not work on either SL6 or SL7 for users with home directories in AFS. This problem is caused by a security fix, the details are in the Redhat bug #697485. A patch has been created to fix the problem in SL6, it will go out in the next stable release after which we will inform the users of the problem. It would be useful to know how many people were affected, Stephen will check the logs. We will need to patch the SL7 version at some point.

There is a new major release of the SL6 devtoolset. This brings it into the main software collections repository which is a good thing but does cause some short-term pain. We will need to warn all users about potential breakage when this goes out next year. We should take the chance to also properly advertise the facility.

Chris has been having problems with creating a third LVM pool using the new disks

hare and wildcat
It turns out that these two machines do not have RAID cards so we can drop that header (and thus silence nagios)

Stephen silenced a load of autoreboot messages from cron on the hadoop cluster about not being able to use the run method because the component is not started.

RT tickets
Still a few more RT tickets to be cleared. Stephen will add #68529 to the MPU activities list and chase up about the new disk for theia. Chris will look into whether the problem with the Optiplex 745 can be fixed by replacing the machine with something a bit newer.


flash player
Thanks to Iain Rae we have a package for the latest flash player which fixes the problems with firefox now blocking old versions. It will go out in the next stable release.

This Week

  • Alastair
    • systemd project
      • Consider how components will work with systemd
      • Continue work on documentation - guidance for other COs on how to use
      • Look at getting Stop method to rebuild /etc/systemd regardless of whether there have been resource changes (remember Stop doesn't call Configure)
      • convert to module
    • EL7 project
      • what sort of level of space is required by systemd journald logging (for desktop /var sizing)
        • (By default journald logs to /run/log. Have to mkdir /var/log/journal to keep data). Have enabled on one machine
        • identify default retention policyDefault retention is to use up to 10% of partition. Can use either space or time as a constraint on space. Logs are per user + system, so users can read their own data. Each log file starts at 8MB, so a popular machine will have lots of log data.
        • Blog about journald retention policy - and document how to set...
        • Blog about decision to keep journald and /var/lcfg/log/syslog duplication - and resulting configuration change.
      • check installroot stuff same version across SL6 and EL7
        • and pull out old SL5 stuff
      • Look at whether we need anything better than existing network component for desktops - VirtualBox works fine.
      • Look at lightdm issues
        • locking (dm-lock doesn't actually lock, just respawns greeter on different VT - can still switch back to original VT
        • power management
      • Blog about dconf and lightdm components
      • Look at LCFG bug #799 (systemd buffered output)
      • convert lcfg-dconf and lcfg-lightdm to module
      • Fix systemd, dconf, and lightdm components so appropriate default values are in the defaults file, not just in the schema file
      • Check dconf and lightdm schemas at LCFG level - should be same as live/defaults.h. Check dconf in LCFG default desktop
      • Take to CEG - DICE EL7 by 1st Feb, COs desktops in February, guinea pigs by 1st March.
    • RT 65774 - try two identical monitors on my machine
    • Need to remove default bridge from kvmtool create
    • Think about disk partition policy
    • Review last reviewed date for documentation
    • Consider more cores as default for KVM guests
    • Is there a way of disabling debugging information being displayed by drupal when there are problems?Can't see how to do safely (needs disable backtrace in /etc/php.ini?)- Ask David Marsh in Physics?
    • Read LISA notes
    • Look at KVM server loading
    • At some point - look at installroot kdcregister solution
    • Look at lcfg-lvm component BUG #807)
    • Look at jubilee lvm config Design flaw in lvm component - requires scsi /dev names to be static, which on a SAN attached machine isn't the case. Surprised haven't tripped up already on this

  • Chris
    • EL7
      • continue with Sleep component
      • investigate Gnome power-management and document
    • url shortener (once gdm solved)
    • Create Project entries - for KVM refinement project
    • Experiment rename br0 as br33 on metropolitan
    • Think about disk partition policy
    • Review last reviewed date for documentation
    • Commission jubilee's extra disks as an extra pool
    • Identify VMs to move to waterloo and move to balance load
    • Quick look at rootmail
    • RT 69276, 61762

  • Stephen
    • LCFG client refactor stage 1
      • schedule debrief meeting
    • EL7
      • Continue thinking about functionality
      • Complete porting MPU managed resources to the DICE level
        • Coordinate a meeting of COs - for first week of January
      • Finish working towards a stable EL7 release
    • Think about PD - Interested in ZeroMQ
    • Deploy northern as staff.nx (first open up holes and test from home) - awaiting disks
    • Think about disk partition policy
    • Review last reviewed date for documentation
    • Add extra memory to waterloo (and if those work, order up more memory for hammersmith)
    • Incorporate comments into pandemic security doc
    • Blog/announce cronie patch to users
    • RT 68529, 68269

-- AlastairScobie - 16 Dec 2014

Topic revision: r9 - 13 Jan 2015 - 14:47:37 - AlastairScobie
This site is powered by the TWiki collaboration platformCopyright © by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding TWiki? Send feedback
This Wiki uses Cookies