MPU Meeting Tuesday 7th January 2014

Inventory

Nothing happened.

Virtual DICE

Toby has investigated what LDAP access is required for virtual DICE to work on external networks. A meeting of interested parties is being organised. After the meeting we can decide what further work will be done and what should be held over for the next virtual DICE project. Chris noted that we said that we would make new images before the start of semester 2 to pickup any new teaching software.

LCFG Client Refactoring

Stephen now has a working prototype system for processing the XML profiles and creating DBM and rpmcfg files. This works quite nicely but requires a lot of memoery (approximately 100MB at peak), mostly that is due to Moose which is very heavy. Stephen is now going to look at the more lightweight OOP "Moo" framework to see if it can give us what we want without the large memory overhead.

LCFG Git Feasibility

Stalled for now.

Miscellaneous Development

Alastair has been looking at systemd in preparation for SL7.

Operational

computing.help
Alastair has added a new virtual disk for the computing.help service and mounted it as /var so that it won't run out of disk space. Currently the old partition is still there, at some point it needs to be removed.

libvirt
At some point we should consider using the latest libvirt packages on SL6.

rfe archive command
Chris will add the rfe archive command provided by Graham for the LCFG profiles.

podium machines and shutdown
The idea of preventing of users from shutting down podium machines was discussed. It was clear that this cannot be done with an access.conf file as gnome issues the shutdown command via dbus.

disk full warnings
Chris will look at adding a cron job to send simple warnings to root mail when the disk usage for specified partitions goes over a specified limit. This is to help spot desktop machines which have full root partitions where we won't want to do the monitoring with nagios.

diceservices and MPU
We need to check the diceservices list to see if any MPU services are missing.

This Week

  • Alastair
    • Start Inventory project diary
    • Inventory project
      • Submit bug/enh to App::Cmd author wrt option to die on unspecified options
      • Pester George about location API
      • Publish revised schema (and document)
      • Continue fleshing out processes (by end of 2013)
    • Order a spare 600GB disk for waterloo (hot spare)
    • Ask George - what does the TXretransmit value mean for switch connections? (Can't see this using rfe -xl )
    • Consider how to make metropolitan usable by users
      • Decide whether we do this now or not
      • ISOs
      • minimal docs (mostly manual)
      • they'll use virt-manager, but not create machines or change config
    • circulate table of LCFG bugs
    • Consider activities list
    • Look at gnome issue with NX - try inftest account
    • Finish off RT tickets
    • Read LCFG XML documentationPointed out no mention of nested RECORDs
    • Look to see if there is gconf option to remove "shutdown machine" from menu item (quick look)
    • Accumulate effort for all running projects - just use normal figures spreadsheet for this
    • Delete old /var on hilfe
    • Look at adding wake.inf, to security scanAsked Ian to add to the next batch of new sites to be scanned
    • Send scan result for www.lcfg.org to Stephen
    • Look at updaterpms status check email again
      • status check checks whether there has been a successful update within the last 24 hours. However, whilst there is an enforced status upload when a machine wakes, there is no such upload when a machine sleeps. I think what's happening is that machines sleep before the client gets round to reporting the new status to the LCFG servers.

  • Chris
    • Tidy up RT tickets as per discussion 17/12
    • Make new virtual DICE images
    • Organise server firmware checks more (document and show Carol what to do)
    • Chris to look at implementing Graham's suggestions on rfe archiving an LCFG profile
    • Look at producing a very simple cron script to flag specified partitions over 90% (parameter). mail to rootmail
    • consider new services for diceservices

  • Stephen
    • LCFG client refactor stage 1
      • report
      • schedule debrief meeting for new year
    • LCFG client refactor stage 2
      • continue working on mock up
    • Consider dhcpd component changes
    • Consider activities list
    • Produce wiki page listing what needs changing when archiving an OS platform
    • Run kernel version check script (so we can find which of our servers are still running old kernels)
    • Check with SEE what they did to improve NX performance
    • close off RT #64886
    • consider new services for diceservices

  • Carol

-- AlastairScobie - 07 Jan 2014

Topic revision: r5 - 14 Jan 2014 - 11:30:11 - ChrisCooke
 
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