[Planetlab-devel] latest 4.2 issues

[Stephen Soltesz]

I think we're very close to having a complete 4.2 release.

The only parts missing from the official 4.2 release are the db-config, 
traceroute fix, pl_netflow initscript.  The rest is administrative on our side 
wrt the actual deployment.

Stephen.

TODO for 4.2 release:
  * Will 4.1 nodes continue to work with 4.2 PLC?  Updating PLC to 4.2. (Tony)
  * Sync 4.2 build repo for alpha to prevent yum problems w/ 4.1 repo (Marc)
  * db-config should set the default confFile values for yum.conf plc repo (Marc)
  * pl_netflow initscript with timeout, to work around vsys/nm race. (Faiyaz)
  * traceroute reports too much info (such as coblitz traffic) to slices. (Sapan)
  * Keep an eye out for COW errors. (Stephen)

On-going:
  * Continue developing pdelta-GUI (Sapan)
  * Roll-out tools (Stephen)
  * Continue developing pdelta (Stephen)

Done/Postponed:
  * ping issue (Sapan)
  * NM leaves slices with 'defunct' python processes. (Faiyaz)
  * Tag kernel with latest config options and networking fixes.
  * yum.conf in the root context needs to work by default (Marc)
  * planetflow needs a landing page until we have the ready-for-live version. 
(Stephen)
  * PLC_PLANETFLOW_HOST gets propogated to alpha nodes.  Ready for landing page.
  * Find/fix reboot/halt problem on alpha nodes.  Related to failed vserver 
shutdowns. (Stephen)
  * fix to audit_log_user_command() (Daniel)
  * re-enable cow support in kernel.  not in RC3. (Stephen)
  * Clarify changes to Scott Baker for removed copy-on-write support in Stork 
(todo at PL Dev workshop)
  * pl_netflow correct setup: run pfmount on init, access /var/local/fprobe. 
(Stephen)
  * testing netflow with slice_ids (Stephen)