Editing Containers/Mini-summit 2008

Jump to: navigation, search

Warning: You are not logged in. Your IP address will be publicly visible if you make any edits. If you log in or create an account, your edits will be attributed to your username, along with other benefits.

The edit can be undone. Please check the comparison below to verify that this is what you want to do, and then save the changes below to finish undoing the edit.
Latest revision Your text
Line 1: Line 1:
There was a containers mini-summit at the [http://www.linuxsymposium.org/2008/ OLS'08]. This page served for organizing this event.
+
There will be a containers mini-summit at the [http://www.linuxsymposium.org/2008/ OLS'08]. This page is for organizing this mini-summit. Feel free to edit.
  
'''When''': 22nd of July 2008, 8:30-16:30<br/>
+
'''When''': 22nd of July 2008<br/>
'''Where''': Ottawa, ON, Canada, Novotel Hotel (Albion A).
+
'''Where''': Ottawa, ON, Canada.
 
 
== Notes ==
 
 
 
For the notes from the mini-summit, '''see [[Containers/Mini-summit 2008 notes]].'''
 
  
 
== Proposal ==
 
== Proposal ==
  
The mini-summit proposal sent to OLS organizers; '''see [[/Proposal|proposal]]'''.
+
The mini-summit proposal sent to OLS organizers. '''See [[/Proposal|proposal]]'''.
  
 
== Topics to discuss ==
 
== Topics to discuss ==
Line 28: Line 24:
 
<!-- Put this in three columns if browser is smart enough -->
 
<!-- Put this in three columns if browser is smart enough -->
 
<div style="-moz-column-count:3; -webkit-column-count:3; column-count:3; text-align: left; background: #fefef0; border: 1px solid #ddddc0;">
 
<div style="-moz-column-count:3; -webkit-column-count:3; column-count:3; text-align: left; background: #fefef0; border: 1px solid #ddddc0;">
 +
# Kir Kolyshkin
 
# Pavel Emelyanov
 
# Pavel Emelyanov
 
# Denis Lunev
 
# Denis Lunev
Line 33: Line 30:
 
# Serge Hallyn
 
# Serge Hallyn
 
# Dave Hansen
 
# Dave Hansen
 +
# Cedric Le Goater
 
# Daniel Lezcano
 
# Daniel Lezcano
 
# Srivatsa Vaddagiri
 
# Srivatsa Vaddagiri
Line 48: Line 46:
 
# Bart Trojanowski
 
# Bart Trojanowski
 
# Joseph Ruscio
 
# Joseph Ruscio
# Constant Chan
 
# Linda Knippers
 
# Satoshi Uchida
 
# Masahiko Takahashi
 
# Martine Silbermann
 
# Benoit des Ligneris
 
# Patrick Naubert
 
# Daisuke Nishimura
 
# Sudhir Kumar
 
# Munehiro Ikeda
 
# Kamalesh Babulal
 
# John Schulz
 
# Poornima Nayak
 
# Gyuil Cha
 
# YoungHo Kim
 
# Rob Woolley
 
# Daniel Robbins
 
# Jason Baron
 
# Subrata Modak
 
# Veerendra C
 
# Joe MacDonald
 
# Andrew Theurer
 
# Myron Stowe
 
# Peter Teoh
 
# Ricky Liang
 
 
</div>
 
</div>
  
 
== Agenda ==
 
== Agenda ==
  
* Namespaces/Containers (8:30am-11am)
+
* Namespaces/Containers
** sysfs issues (and any /proc issues)
 
*** uevents/hotplug
 
** Network namespaces issues
 
*** multiple namespaces in one process
 
** Device namespace design?
 
** User namespace
 
 
** Additional needed namespaces
 
** Additional needed namespaces
 
*** Small namespaces ''What to do with small subsystem that might need virtualization. E.g. in openvz we have FUSE, binfmt_misc and some other small stuff virtualized. But how to merge it in mainline? Create a separate namespace for each? Mere them into one? How to call this then?''
 
*** Small namespaces ''What to do with small subsystem that might need virtualization. E.g. in openvz we have FUSE, binfmt_misc and some other small stuff virtualized. But how to merge it in mainline? Create a separate namespace for each? Mere them into one? How to call this then?''
 +
** Nature of a 'container' — kernel object or userspace fiction
 +
** Handling of /proc and /sysfs within containers
 
** Handling filesystem/namespace synchronization  (not sure what the issue is)
 
** Handling filesystem/namespace synchronization  (not sure what the issue is)
** Container design
+
** How to enter a container
*** How to enter a container
+
** User namespaces?
*** Nature of a 'container' — kernel object or userspace fiction
+
* Cgroups+Resource management
 
 
* Cgroups+Resource management (11:30-2pm)
 
 
** Cgroup implementation
 
** Cgroup implementation
*** Locking (don't let cgroup_lock() become the BKL)
 
*** Transactional attachment
 
*** "procs" file
 
*** User-space notification API
 
**** Resource counter hit soft/hard limit
 
**** Task entered/left cgroup
 
**** OOM occurred
 
*** Binary statistics API
 
** Existing cgroups
 
*** Memory (Balbir's NOTE: I would prefer to take some of this discussion to my memory controller BoF on Wednesday. Lets discuss this at the end)
 
**** Supporting over-commit and guarantees
 
**** Soft-limits
 
**** Hierarchical borrowing - in kernel or userspace?
 
**** Per-cgroup refault information?
 
*** Kernel memory
 
*** Device
 
*** Memrlimit
 
**** Some push-back over this - can we give real use cases?
 
*** CPU scheduler
 
 
** Additional cgroups and their design
 
** Additional cgroups and their design
*** Swap (separate subsystem or merge with memory?)
+
** libcg - userspace explotation of control                           groups/resource management
*** Disk I/O (several proposed designs)
+
** Resource management
*** Network traffic classification
+
* Checkpoint/Restart
*** Freezer
+
** Summary of existing c/r patchsets/designs
*** Signaller
+
** How to initiate and synchronize checkpoint/restart
*** OOM Handler
+
** (state of freezer subsystem?)
** libcg - userspace explotation of control groups/resource management
+
** Memory state dump
*** Overview so far
+
** How to dump/fetch data for resource (file, ipc) checkpoint
*** Is kernel-based reclassification needed?
+
** How to do restart
*** Real use-cases
+
** Hopefully we can make decisions here, and get a bit of a hackfest going during OLS
*** Future directions
 
 
 
 
 
* Checkpoint/Restart (2:30pm-5pm)
 
** Documentation : Look at "See Also" section below
 
** Goals and expectations of this summit
 
*** identify, discuss and (if possible) agree on the general design
 
*** identify, discuss and (if possible) agree on the technical points
 
*** decide on priorities for different components (eg. high, medium, low)  such that the final outcome is a practical road-map that would keep us busy for (at least) until the next OLS (though the "O" may change ;)
 
** What are the problems that the linux community can solve with the checkpoint/restart ?
 
** Preparing the kernel internals
 
*** How we implement it without affecting long term maintainability ?
 
*** What are the kernel subsystems, process resources and framework for CR ?
 
*** Which pieces to target first ?
 
 
 
The following technical points can be discussed during the mini-summit if we have time or later at the OLS.
 
 
 
** Checkpointing / Restarting
 
*** Reaching a quescient point - network, processes, aio, avoiding side effects of quiesce/revive
 
*** Checkpoint - signal handler ? syscall ? crfs ? process hierarchy, resource dependencies, system and process resources
 
*** Restarting - New binary format handler ? converting between formats (from older kernel to newer)
 
*** Notification to processes which explicitly wish to be notified about quiesce, checkpoint and restart - container state ? new signals ?
 
** Determining the userspace API - Posix 1003.1m ?
 
** Passing the kernel internal state to/from userspace - coredump like file ? swap per container ? netlinks, CR filesystem ? army of different call for the CR (proc, existing syscalls, ...)
 
** Hopefully we can continue to discuss in the next days and get a bit of a hackfest going during OLS :)
 
  
 
== Moderators ==
 
== Moderators ==
Line 153: Line 76:
 
* Namespaces/containers: Serge Hallyn, Dave Hansen
 
* Namespaces/containers: Serge Hallyn, Dave Hansen
 
* Cgroups and resource management: Paul Menage, Balbir Singh, Dhaval Giani
 
* Cgroups and resource management: Paul Menage, Balbir Singh, Dhaval Giani
* Checkpoint/restart: Daniel Lezcano, Oren Laadan
+
* Checkpoint/restart: Cedric Le Goater, Oren Laadan
  
 
== See also ==
 
== See also ==
 
* http://www.linuxsymposium.org/2008/cfp.php — OLS call for papers
 
* http://www.linuxsymposium.org/2008/cfp.php — OLS call for papers
 
* https://lists.linux-foundation.org/pipermail/containers/2008-January/009688.html
 
* https://lists.linux-foundation.org/pipermail/containers/2008-January/009688.html
* http://openvz.org/pipermail/devel/2008-July/012891.html
 
* Checkpoint/Restart
 
** Zap : http://www.ncl.cs.columbia.edu/publications/usenix2007_fordist.pdf
 
** Metacluster : http://lxc.sourceforge.net/doc/ols2006/lxc-ols2006.pdf
 
** OpenVZ : [[Checkpointing and live migration]]
 
** Checkpoint/Restart technology : http://en.wikipedia.org/wiki/Application_checkpointing
 
** Virtual Servers and Checkpoint/Restart in Mainstream Linux : Sigops document
 
** Remote fork: http://www.cse.nd.edu/~dthain/courses/classconf/wowsys2004/talks/rfork.pdf
 
** Vmadump : http://bproc.sourceforge.net/c268.html
 
** Posix CR : http://techpubs.sgi.com/library/tpl/cgi-bin/getdoc.cgi/0650/bks/SGI_Admin/CPR_OG/sgi_html/ch03.html
 
** An OS services overview : http://sw-eng.falls-church.va.us/itsg/P08V31.htm
 
  
 
[[Category: Containers]]
 
[[Category: Containers]]
 
[[Category: Events]]
 
[[Category: Events]]

Please note that all contributions to OpenVZ Virtuozzo Containers Wiki may be edited, altered, or removed by other contributors. If you don't want your writing to be edited mercilessly, then don't submit it here.
If you are going to add external links to an article, read the External links policy first!

To edit this page, please answer the question that appears below (more info):

Cancel Editing help (opens in new window)