On 02/10/2011 01:17 AM, Sukadev Bhattiprolu wrote:
Oren Laadan [or...@cs.columbia.edu] wrote:
|
| To address this, initially I thought that we could make ghosts take
| the tasklist_lock (write) when they change their -exit_signal.
|
| But that's not enough because the parent may already
On 02/09/2011 07:01 AM, Louis Rilling wrote:
On 08/02/11 18:09 -0800, Sukadev Bhattiprolu wrote:
Oren Laadan [or...@cs.columbia.edu] wrote:
|
|
| On 02/05/2011 04:40 PM, Sukadev Bhattiprolu wrote:
| Oren Laadan [or...@cs.columbia.edu] wrote:
| | Suka,
| |
| | This patch
On 02/09/2011 09:44 PM, Sukadev Bhattiprolu wrote:
Oren Laadan [or...@cs.columbia.edu] wrote:
|
|
| But if that is true, I need to investigate why Louis Rilling was getting
| the crash in Jun 2010 - which he tried to fix here:
|
|http://lkml.org/lkml/2010/6/16/295
|
| I see
On 02/08/2011 09:09 PM, Sukadev Bhattiprolu wrote:
Oren Laadan [or...@cs.columbia.edu] wrote:
|
|
| On 02/05/2011 04:40 PM, Sukadev Bhattiprolu wrote:
| Oren Laadan [or...@cs.columbia.edu] wrote:
| | Suka,
| |
| | This patch - and the corresponding kernel patch - are wrong
Hi,
This patch-set changes the way pids are saved/restored to treat them
as proper shared objects (tracked in objhash) rather than simple pid
numbers. The patch-set applies on top of branch 'ckpt-v23-rc1'. It
also requires updates to user-cr (patchset posted separately).
Changelog[v2]:
- Wipe
This helpoer is Useful to get the pid from the root task's (checkpoint
or restart) point of view.
Signed-off-by: Oren Laadan or...@cs.columbia.edu
---
include/linux/checkpoint.h | 10 +-
kernel/checkpoint/checkpoint.c |4 ++--
kernel/checkpoint/restart.c|6
In preparation for next (super)-patch.
Signed-off-by: Oren Laadan or...@cs.columbia.edu
---
restart.c | 44 ++--
1 files changed, 22 insertions(+), 22 deletions(-)
diff --git a/restart.c b/restart.c
index 05d101b..966f7a1 100644
--- a/restart.c
+++ b
Signed-off-by: Oren Laadan or...@cs.columbia.edu
---
restart.c | 22 ++
1 files changed, 10 insertions(+), 12 deletions(-)
diff --git a/restart.c b/restart.c
index 8106fd6..195a892 100644
--- a/restart.c
+++ b/restart.c
@@ -359,35 +359,32 @@ static void sigint_handler(int
if ctx-errno isn't already set, the:
- ctx_set_errno() saved errno in ctx-error
- ctx_ret_errno() sets ctx-error (and returns -1)
Signed-off-by: Oren Laadan or...@cs.columbia.edu
---
common.h |3 ++
restart.c | 99
2 files
terinating, and makes the coordiantor
collect the feeder.
Cc: Sukadev Bhattiprolu suka...@linux.vnet.ibm.com
Signed-off-by: Oren Laadan or...@cs.columbia.edu
---
restart.c | 86 ++--
1 files changed, 49 insertions(+), 37 deletions(-)
diff --git
This patchset adds the necessary support in user-cr related to
handling of pids as proper shared objets. You must use this if you use
the corresponding kernel-cr patchset recetly posted.
Changelog[v2]:
- Cleanups and many bug fixes: it now passes all tests in tests-cr
Oren.
Signed-off-by: Oren Laadan or...@cs.columbia.edu
---
restart.c | 19 +--
1 files changed, 17 insertions(+), 2 deletions(-)
diff --git a/restart.c b/restart.c
index a1af631..05d101b 100644
--- a/restart.c
+++ b/restart.c
@@ -158,6 +158,9 @@ struct ckpt_ctx {
struct
---
include/linux/checkpoint_hdr.h | 29 -
1 files changed, 24 insertions(+), 5 deletions(-)
diff --git a/include/linux/checkpoint_hdr.h b/include/linux/checkpoint_hdr.h
index f7c4d9a..227bfbe 100644
--- a/include/linux/checkpoint_hdr.h
+++
.
Signed-off-by: Oren Laadan or...@cs.columbia.edu
---
restart.c | 50 +-
1 files changed, 45 insertions(+), 5 deletions(-)
diff --git a/restart.c b/restart.c
index 966f7a1..01566c2 100644
--- a/restart.c
+++ b/restart.c
@@ -306,6 +306,18 @@ static
the dead
pgids and the tgids.
The patch also refactors ckpt_set_task() into smaller pieces to
simplify its logic and improve readability.
Signed-off-by: Oren Laadan or...@cs.columbia.edu
---
restart.c | 83 ++--
1 files changed, 63
Signed-off-by: Oren Laadan or...@cs.columbia.edu
---
kernel/signal.c |7 +--
1 files changed, 5 insertions(+), 2 deletions(-)
diff --git a/kernel/signal.c b/kernel/signal.c
index dca40be..3842f5d 100644
--- a/kernel/signal.c
+++ b/kernel/signal.c
@@ -2871,7 +2871,7 @@ static void
objects are inserted first, so we will
need to know how many objects were inserted earlier - so we will alos
save the count at that time.
Signed-off-by: Oren Laadan or...@cs.columbia.edu
---
include/linux/checkpoint.h |2 ++
kernel/checkpoint/objhash.c |5 +
2 files changed, 7
Add the interface to handle pids as proper c/r shared objects. This
is the first step in converting the c/r code to hanlde pid as pids
objects that are namespace aware. Also remote the unused field
@root_pid from struct ckpt_ctx.
Signed-off-by: Oren Laadan or...@cs.columbia.edu
---
include
, and in userspace we'll
need to update the logic of MakeForest algorithm to be pid-ns aware
(probably similarly to how sid constraints are handled).
Signed-off-by: Oren Laadan or...@cs.columbia.edu
---
include/linux/checkpoint_hdr.h | 23 ++-
include/linux/checkpoint_types.h | 10 +-
kernel/checkpoint
pre-deceases a
ghost task.
Is this still a problem in 2.6.37 ?
Oren.
On 01/10/2011 08:51 PM, Oren Laadan wrote:
Applied to user-cr.
Thanks,
Oren.
On 12/10/2010 10:35 PM, Sukadev Bhattiprolu wrote:
From: Sukadev Bhattiprolu suka...@linux.vnet.ibm.com
Date: Fri, 10 Dec 2010 19:23:58
On 02/05/2011 04:40 PM, Sukadev Bhattiprolu wrote:
Oren Laadan [or...@cs.columbia.edu] wrote:
| Suka,
|
| This patch - and the corresponding kernel patch - are wrong
Ah, I see that now.
But am not sure about the kernel part though. We were getting a crash
reliably (with older kernels
Suka,
Thanks for the review.
On 02/05/2011 04:43 PM, Sukadev Bhattiprolu wrote:
Oren:
I am still reviewing this patchset, but have a few questions/comments
below on this patch.
| From: Oren Laadan or...@cs.columbia.edu
| Subject: [PATCH 4/5] c/r: checkpoint and restart pids objects
On 02/05/2011 05:02 PM, Oren Laadan wrote:
On 02/05/2011 04:40 PM, Sukadev Bhattiprolu wrote:
Oren Laadan [or...@cs.columbia.edu] wrote:
| Suka,
|
| This patch - and the corresponding kernel patch - are wrong
Ah, I see that now.
But am not sure about the kernel part though. We were
[...]
+file = fget(fds[1-which]);
+if (!file) /* this should _never_ happen ! */
If it should never happen then this should be:
BUG_ON(!file);
that way we see the bug as soon as it's detected rather than having to
dig through the stack and
Thanks for testing - pushed both patches to v23-rc1.
Oren.
On 02/01/2011 02:38 PM, Dan Smith wrote:
OL Original patch posted by Dan Smith.
Tested-by: Dan Smith da...@us.ibm.com
With this and my pipe refcount fix, all my pipe tests pass, including
the ones that were stuck before because
Thanks, pushed to v23-rc1.
Oren
On 02/01/2011 01:47 PM, Dan Smith wrote:
Oren's version of my patch leaks pipe objects due to the way the fget() and
fput() operations are structured. The intent was to avoid getting and putting
references more than necessary, but the result was a more
thanks, pushed to v23-rc1.
On 01/31/2011 11:44 AM, Dan Smith wrote:
OL Cc: Dan Smith da...@us.ibm.com
OL Signed-off-by: Oren Laadan or...@cs.columbia.edu
Tested-by: Dan Smith da...@us.ibm.com
___
Containers mailing list
contain...@lists.linux
I modifed the patch a bit according to our IRC chat today:
From 9c74f82411d77cf0194a17ba99af0dd31070e88a Mon Sep 17 00:00:00 2001
From: Oren Laadan or...@cs.columbia.edu
Date: Mon, 31 Jan 2011 19:01:49 -0500
Subject: [PATCH] c/r: clear the objhash before completing restart, but delay
free (v3
On 01/28/2011 03:18 PM, Eric W. Biederman wrote:
Serge E. Hallyn serge.hal...@canonical.com writes:
Hi,
Now that we are allowing udev to run in containers, Daniel has
noticed that updates to sysfs uevent files will trigger a flurry
of activity in all containers on the host. While not a
Dan,
Thanks for pointing this out and the patch.
I think it would be simpler to use ckpt_kread(), no ?
If so, I'll go ahead and fix then import.
Oren.
On 01/28/2011 01:45 PM, Dan Smith wrote:
While this fixes restoring pipes that were completely full, it actually
corrects a potential issue
On 01/28/2011 06:08 PM, Matt Helsley wrote:
On Fri, Jan 28, 2011 at 10:45:11AM -0800, Dan Smith wrote:
While this fixes restoring pipes that were completely full, it actually
corrects a potential issue with restoring any pipe buffers. By using
splice() to do this work when we are reading
From 7d3f62995ed4e83bb43deafa362c78624099e495 Mon Sep 17 00:00:00 2001
From: Oren Laadan or...@cs.columbia.edu
Date: Fri, 28 Jan 2011 19:23:17 -0500
Subject: [PATCH] c/r: fix restoring pipes with full buffers
Dan Smith pointed out a problem with the use of splice to restore a
pipe's contents
Hi,
This patch-set changes the way pids are saved/restored to treat them
as proper shared objects (tracked in objhash) rather than simple pid
numbers. The patch-set applies on top of branch 'ckpt-v23-rc1'. It
also requires updates to user-cr (patchset posted separately).
Thanks,
Oren.
Signed-off-by: Oren Laadan or...@cs.columbia.edu
---
kernel/signal.c |7 +--
1 files changed, 5 insertions(+), 2 deletions(-)
diff --git a/kernel/signal.c b/kernel/signal.c
index dca40be..3842f5d 100644
--- a/kernel/signal.c
+++ b/kernel/signal.c
@@ -2871,7 +2871,7 @@ static void
Add the interface to handle pids as proper c/r shared objects. This
is the first step in converting the c/r code to hanlde pid as pids
objects that are namespace aware. Also remote the unused field
@root_pid from struct ckpt_ctx.
Signed-off-by: Oren Laadan or...@cs.columbia.edu
---
include
This helpoer is Useful to get the pid from the root task's (checkpoint
or restart) point of view.
Signed-off-by: Oren Laadan or...@cs.columbia.edu
---
include/linux/checkpoint.h | 10 +-
kernel/checkpoint/checkpoint.c |5 +++--
kernel/checkpoint/restart.c|6
, and in userspace we'll
need to update the logic of MakeForest algorithm to be pid-ns aware
(probably similarly to how sid constraints are handled).
Signed-off-by: Oren Laadan or...@cs.columbia.edu
---
include/linux/checkpoint_hdr.h | 22 ++-
include/linux/checkpoint_types.h | 10 +-
kernel/checkpoint
Make tty_old_pgrp and tty_pgrp use the shared pid instead of saving
the actual pid number in {checkpoint,restore}_signal().
Signed-off-by: Oren Laadan or...@cs.columbia.edu
---
kernel/signal.c | 42 --
1 files changed, 32 insertions(+), 10 deletions
/2011 11:10 AM, Oren Laadan wrote:
This helpoer is Useful to get the pid from the root task's (checkpoint
or restart) point of view.
Signed-off-by: Oren Laadan or...@cs.columbia.edu
---
include/linux/checkpoint.h | 10 +-
kernel/checkpoint/checkpoint.c |5 +++--
kernel
Hi,
This patchset adds the necessary support in user-cr related to
handling of pids as proper shared objets. You must use this if you use
the corresponding kernel-cr patchset recetly posted.
Thanks,
Oren.
___
Containers mailing list
if ctx-errno isn't already set, the:
- ctx_set_errno() saved errno in ctx-error
- ctx_ret_errno() sets ctx-error (and returns -1)
Signed-off-by: Oren Laadan or...@cs.columbia.edu
---
common.h |3 ++
restart.c | 99
2 files
Signed-off-by: Oren Laadan or...@cs.columbia.edu
---
restart.c | 22 ++
1 files changed, 10 insertions(+), 12 deletions(-)
diff --git a/restart.c b/restart.c
index 8106fd6..195a892 100644
--- a/restart.c
+++ b/restart.c
@@ -359,35 +359,32 @@ static void sigint_handler(int
The initialization belongs to the caller of the cr-library - i.e.
restart-main and checkpoint-main, and not in the cr-library.
Signed-off-by: Oren Laadan or...@cs.columbia.edu
---
checkpoint-main.c |4 ++--
checkpoint.c |8
restart-main.c|4 ++--
3 files changed, 4
---
include/linux/checkpoint_hdr.h | 28 +++-
1 files changed, 23 insertions(+), 5 deletions(-)
diff --git a/include/linux/checkpoint_hdr.h b/include/linux/checkpoint_hdr.h
index f7c4d9a..8087250 100644
--- a/include/linux/checkpoint_hdr.h
+++
Signed-off-by: Oren Laadan or...@cs.columbia.edu
---
restart.c | 19 +--
1 files changed, 17 insertions(+), 2 deletions(-)
diff --git a/restart.c b/restart.c
index 9535543..d998834 100644
--- a/restart.c
+++ b/restart.c
@@ -157,6 +157,9 @@ struct ckpt_ctx {
struct
In preparation for next (super)-patch.
Signed-off-by: Oren Laadan or...@cs.columbia.edu
---
restart.c | 44 ++--
1 files changed, 22 insertions(+), 22 deletions(-)
diff --git a/restart.c b/restart.c
index d998834..ca02383 100644
--- a/restart.c
+++ b
Signed-off-by: Oren Laadan or...@cs.columbia.edu
---
ckptinfo.c |4 ++--
1 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/ckptinfo.c b/ckptinfo.c
index d73b38c..1361c21 100644
--- a/ckptinfo.c
+++ b/ckptinfo.c
@@ -254,7 +254,7 @@ static int image_parse(int fd, struct args *args
Note that you also need to apply the corresponding user-cr patchset
to sync the userspace tools with this new logic.
Oren.
On 01/26/2011 11:20 AM, Oren Laadan wrote:
Hmmm.. the cover post did not make it :(
This patch-set changes the way pids are handled in c/r - instead
of simple pid
:19 AM, Oren Laadan wrote:
The initialization belongs to the caller of the cr-library - i.e.
restart-main and checkpoint-main, and not in the cr-library.
Signed-off-by: Oren Laadan or...@cs.columbia.edu
---
checkpoint-main.c |4 ++--
checkpoint.c |8
restart-main.c
Hi,
I vaguely recall some discussions/ideas about the possibility
of a devices namespace, its pros and cons, and alternative.
Related to that is also device viruatlization, and isolation
of devices in containers.
Any thoughts and or pointers to past/current discussions are
welcome :)
Thanks,
terinating, and makes the coordiantor
collect the feeder.
Cc: Sukadev Bhattiprolu suka...@linux.vnet.ibm.com
Signed-off-by: Oren Laadan or...@cs.columbia.edu
---
restart.c | 86 ++--
1 files changed, 49 insertions(+), 37 deletions(-)
diff --git
On 01/21/2011 02:39 AM, Sukadev Bhattiprolu wrote:
Oren,
In usercr commit 0dd3f9dacc0a08f5fb562e385a4e4244f453b41c we have the
following
diff:
@@ -1868,10 +1902,16 @@ static int ckpt_fork_feeder(struct ckpt_ctx *ctx)
close(ctx-pipe_child[0]);
ctx-pipe_out =
On 01/14/2011 01:59 AM, Sukadev Bhattiprolu wrote:
Oren Laadan [or...@cs.columbia.edu] wrote:
| Folks,
|
| I just pushed out a new v23-rc1 branch of linux-cr. This one is
| rebased to 2.6.37, and contains nearly all the patches pulled
| on v22-dev. I only gave it a brief test drive
On 01/11/2011 07:17 PM, Sukadev Bhattiprolu wrote:
Oren Laadan [or...@cs.columbia.edu] wrote:
|
| The following two patches add c/r support for POSIX file locks. They
| are based on the series posted by Suka some time ago:
|
https://lists.linux-foundation.org/pipermail/containers/2010
Folks,
I just pushed out a new v23-rc1 branch of linux-cr. This one is
rebased to 2.6.37, and contains nearly all the patches pulled
on v22-dev. I only gave it a brief test drive... feel free to
throw all your ammo it.
Rebasing was fairly straightforward, but I'd feel more comfortable
with
The following two patches add c/r support for POSIX file locks. They
are based on the series posted by Suka some time ago:
https://lists.linux-foundation.org/pipermail/containers/2010-October/025855.html
Suka:
When I read your post I thought that boht posix and flock were
supported, but
-off-by: Sukadev Bhattiprolu suka...@linux.vnet.ibm.com
Acked-by: Oren Laadan or...@cs.columbia.edu
---
fs/locks.c | 76 +++
include/linux/fs.h |4 +++
2 files changed, 50 insertions(+), 30 deletions(-)
diff --git a/fs/locks.c b/fs
Bhattiprolu suka...@linux.vnet.ibm.com
Signed-off-by: Oren Laadan or...@cs.columbia.edu
---
Documentation/checkpoint/file-system.txt | 59 +++
fs/checkpoint.c | 22 +--
fs/locks.c | 268 ++
include/linux
Pulled and merged in v23-rc1, with a slight change: NCCS is
known to userspace, so no need for CKPT_TTY_NCCS...
Thanks
Oren.
On 12/22/2010 04:15 AM, Sukadev Bhattiprolu wrote:
From: Sukadev Bhattiprolu suka...@linux.vnet.ibm.com
Date: Fri, 10 Dec 2010 18:36:34 -0800
Subject: [PATCH 1/1] Fix
Hi Suka,
Indeed the use of STDIN_FILENO wasn't too friendly to callers
of the library. However, I think the correct fix is different.
So I fixed that, and added a few other cleanups in an attempt
to ensure that the restart code cleans up properly after being
called (succesfully or unsuccesfully)
Applied to user-cr.
Thanks,
Oren.
On 12/10/2010 10:35 PM, Sukadev Bhattiprolu wrote:
From: Sukadev Bhattiprolu suka...@linux.vnet.ibm.com
Date: Fri, 10 Dec 2010 19:23:58 -0800
Subject: [PATCH 1/1] Ghost tasks must be detached
Ghost processes are created only to help restore orphaned
Merged into v23-rc1.
Thanks,
Oren.
On 12/10/2010 10:33 PM, Sukadev Bhattiprolu wrote:
Oren,
As discussed (a while ago)
https://lists.linux-foundation.org/pipermail/containers/2010-November/025905.html
here is a patch. I will send a follow-on user-cr patch to mark
the task detached
Hi,
I applied this changes (a modified version) as part of the
cleanup I mentioned to user-cr a few minutes ago ...
Thanks,
Oren.
On 04/09/2010 11:25 PM, Sukadev Bhattiprolu wrote:
From: Sukadev Bhattiprolu suka...@linux.vnet.ibm.com
Date: Fri, 9 Apr 2010 18:36:46 -0700
Subject:
At the moment, at least @klogd field has to be set manually.
Perhaps it would be better to provide a pre-initialized struct,
so that users could do:
args = CR_RESTART_ARGS;
or
cr_init_restart_args(args);
?
Oren
On 02/24/2010 10:47 PM, Sukadev Bhattiprolu wrote:
global_send_sigint is
This patch prepares the ground for a smooth integration of the support
for hugetlb page coming in following patches.
Cc: Nathan Lynch n...@pobox.com
Signed-off-by: Oren Laadan or...@cs.columbia.edu
---
include/linux/checkpoint.h |4 ++--
ipc/checkpoint_shm.c |4 ++--
mm
.
This depends on the patch support ipc shm regions which are partially
mapped posted to the containers list on 6 August.
Signed-off-by: Nathan Lynch n...@pobox.com
Acked-by: Oren Laadan or...@cs.columbia.edu
---
include/linux/shm.h | 15 ++-
ipc/shm.c | 17 +
2 files
CKPT_VMA_NOT_SUPPORTED
Cc: Nathan Lynch n...@pobox.com
Signed-off-by: Oren Laadan or...@cs.columbia.edu
---
include/linux/checkpoint.h |3 +-
include/linux/checkpoint_hdr.h | 16
include/linux/hugetlb.h| 34 +
ipc/checkpoint_shm.c |2 +-
mm/checkpoint.c
The following 4 patches add supoprt for c/r of HUGETLB pages.
They are based on the original post by Nathan Lynch:
https://lists.linux-foundation.org/pipermail/containers/2010-October/025794.html
Nathan:
I did some serious refactoring of your patches to be able to re-use
the existing
the ipc_shm
checkpoint header after unmapping.
Original patch by Nathan Lynch n...@pobox.com.
Changelog[v23-rc1]:
- Mofidied to reuse existing code in mm/checkpoint.c (specifically
checkpoint_memory_contents() and restore_memory_contents()
Cc: Nathan Lynch n...@pobox.com
Signed-off-by: Oren
Hi,
Thanks for bringing up the issue.
We clearly will need device-specific vma types (in plural!),
and I suspect that the way it will work is that devices will
register their vma types at boot time or dynamically (for
kernel modules).
On 10/20/2010 02:56 PM, Nathan Lynch wrote:
This is
Hi,
I'm not too familiar with all the uses of VM_{IO,RESERVED,PFNMAP}
flags. Should I be concerned about missing out cases where the
checkpoint succeeds but the restart will fail (or worse, succeed
but not work correctly thereafter) ?
If these are only used by devices, then no problem - they
Hi Nathan,
Thanks for the series.
I merged patches 1-3 into v23-rc1.
I posted a reworked version of patches 7-11 for (your) review.
I did not apply 4-6 yet, see my replies to 4,5 first.
Cheers,
Oren
On 10/20/2010 02:56 PM, Nathan Lynch wrote:
Hi Oren,
Here is the hugetlb code rebased to
On 01/10/2011 05:51 PM, Serge Hallyn wrote:
Quoting Bastian Blank (bast...@waldi.eu.org):
On Mon, Jan 10, 2011 at 09:13:34PM +, Serge E. Hallyn wrote:
+ const struct cred *cred = current_cred();
+ const struct cred *tcred = __task_cred(t);
+
+ if (cred-user-user_ns !=
Suka,
Yes, I overlooked it, thanks for pointing out.
The patch looks fine - queued for inclusion.
(I'll remove the now-unneeded define of CKPT_TTY_NCC)
Oren.
On 12/08/2010 08:28 PM, Sukadev Bhattiprolu wrote:
Oren,
Any reason we only checkpoint/restore a 'struct termio' instead of a
'struct
Ok, thanks.
On 12/13/2010 01:27 PM, Sukadev Bhattiprolu wrote:
Oren Laadan [or...@cs.columbia.edu] wrote:
Suka,
Yes, I overlooked it, thanks for pointing out.
The patch looks fine - queued for inclusion.
(I'll remove the now-unneeded define of CKPT_TTY_NCC)
I can post a cleaner patch
On Tue, 23 Nov 2010, Kapil Arya wrote:
OL Even if it did - the question is not how to deal with glue
OL (you demonstrated quite well how to do that with DMTCP), but
OL how should teh basic, core c/r functionality work - which is
OL below, and orthogonal to the glue.
There seems to be an
On Sun, 21 Nov 2010, Gene Cooperman wrote:
As Kapil and I wrote before, we benefited greatly from having talked with
Oren,
and learning some more about the context of the discussion. We were able
to understand better the good technical points that Oren was making.
Since the comparison
On Sat, 20 Nov 2010, Tejun Heo wrote:
Hello,
On 11/20/2010 07:15 PM, Oren Laadan wrote:
[[apologies for the silly prefix on last two posts - a combination
of windows, putty, pine andslow connection is not helping me :( ]]
Maybe it's a good idea to post a clean concatenated version
On Sun, 21 Nov 2010, Gene Cooperman wrote:
Below, we'll summarize the four major questions that we've understood from
this discussion so far. But before doing so, I want to point out that a
single
process or process tree will always have many possible interactions with
the rest of the
On Fri, 19 Nov 2010, Tejun Heo wrote:
Hello,
On 11/19/2010 03:36 PM, Kirill Korotaev wrote:
Can you imagine how many userland APIs are needed to make userspace C/R?
Do you really want APIs in user-space which allow to:
- send signals with siginfo attached (kill() doesn't work...)
Hi,
Based on discussion with Gene, I'd like to clarify key points and
difference between kernel and userspace approaches (specifically
linux-cr and dmtcp): three parts to break the long post...
part I: perpsectice about the types of scopes of c/r in discussion
part II: linux-cr design adn
login as: orenl
Using keyboard-interactive authentication.
Password:
Access denied
Using keyboard-interactive authentication.
Password:
Last login: Fri Nov 19 10:17:21 2010 from 192.117.42.81.static.012.net.il
499:takamine[~]$ pine
PINE 4.64 COMPOSE MESSAGE
login as: orenl
Using keyboard-interactive authentication.
Password:
Access denied
Using keyboard-interactive authentication.
Password:
Last login: Fri Nov 19 10:17:21 2010 from 192.117.42.81.static.012.net.il
499:takamine[~]$ pine
PINE 4.64 COMPOSE MESSAGE
[[apologies for the silly prefix on last two posts - a combination
of windows, putty, pine andslow connection is not helping me :( ]]
___
Containers mailing list
contain...@lists.linux-foundation.org
On 11/17/2010 10:46 AM, Tejun Heo wrote:
Hello, Serge.
On 11/17/2010 04:39 PM, Serge E. Hallyn wrote:
I'm sorry but in-kernel CR already looks like a major misdesign to me.
By this do you mean the very idea of having CR support in the kernel?
Or our design of it in the kernel?
The
On 11/17/2010 05:17 PM, Matt Helsley wrote:
On Wed, Nov 17, 2010 at 12:57:40PM +0100, Tejun Heo wrote:
Hello, Oren.
On 11/07/2010 10:59 PM, Oren Laadan wrote:
snip
Or we could use linux-cr for that: do the c/r in the kernel,
keep the know-how in the kernel, expose (and commit to) a
per
Hi,
Ok, I'll bite the bullet for now - to be continued...
Just one important clarification:
Linux-cr can do live migration - e.g. VDI, move the desktop - in
which case skype's sockets' network stacks are reconstructed,
transparently to both skype (local apps) and the peer (remote apps).
On 11/08/2010 01:37 PM, Gene Cooperman wrote:
Thanks for the careful response, Oren. For others who read this,
one could interpret Oren's rapid post as criticizing the work of
Andres Lagar Cavilla. I'm sure that this was not Oren's intention.
Please read below for a brief clarification of
[cc'ing linux containers mailing list]
On 11/06/2010 04:40 PM, Gene Cooperman wrote:
8. What happens if the DMTCP coordinator ( checkpoint control process) dies?
[ The same thing that happens if a user process dies. We kill the whole
computation, and restart. At restart, we use a
[cc'ing linux containers mailing list]
On 11/07/2010 01:49 PM, Gene Cooperman wrote:
[snip]
Matt had asked how we would handle inotify(), but I was getting swamped
by all the questions. There is a virtualization approach to inotify in which
one puts wrappers around inotify_add_watch(),
Better late than never -
I forgot to CC the list on this thread:
http://lkml.org/lkml/2010/11/2/264
Oren.
___
Containers mailing list
contain...@lists.linux-foundation.org
https://lists.linux-foundation.org/mailman/listinfo/containers
Dan,
Does this patch solve the performance problem in freeing all
the objhash entries upfront (e.g. before letting userspace
resume) ?
If so, is there still a performance hit for doing the 'clear'
portion early before resuming the tasks ? How much does it
depend on the complexity of the
Matt,
This is v2, right ?
If so, is there a changelog ?
Oren.
On 09/23/2010 05:53 PM, Matt Helsley wrote:
This series enables the relinking of unlinked files for the purpose of
enabling checkpoint/restart of tasks with open unlinked files. It
relies on Namhyung Kim's sys-wrapper patches to
On 11/01/2010 01:01 PM, Dan Smith wrote:
OL Does this patch solve the performance problem in freeing all
OL the objhash entries upfront (e.g. before letting userspace
OL resume) ?
OL If so, is there still a performance hit for doing the 'clear'
OL portion early before resuming the tasks ?
On 09/17/2010 04:23 PM, Nathan Lynch wrote:
On Thu, 2010-09-16 at 20:44 -0400, Oren Laadan wrote:
On 09/14/2010 04:02 PM, Nathan Lynch wrote:
Support checkpoint and restore of both private and shared
hugepage-backed mappings established via mmap(MAP_HUGETLB). Introduce
APIs for checkpoint
On 10/06/2010 03:43 PM, Nathan Lynch wrote:
On Thu, 2010-09-16 at 20:37 -0400, Oren Laadan wrote:
Hi Nathan,
Thanks for the patch series. For starters, I'll pull
patches 1-3 for ckpt-v22-dev. I'll wait for reviews
for the others.
Well, reviews haven't exactly been pouring in, but I'm
On 10/30/2010 03:01 AM, Sukadev Bhattiprolu wrote:
From ce9dd2fc7332597d46872f3f8c52ac0806f381d1 Mon Sep 17 00:00:00 2001
From: Sukadev Bhattiprolu suka...@linux.vnet.ibm.com
Date: Fri, 29 Oct 2010 23:16:10 -0700
Subject: [PATCH 1/1] Mark ghost task as detached earlier
During restart() of
Daniel,
Maybe you can throw this on in the series as well ?
http://www.mail-archive.com/linux-...@vger.kernel.org/msg01431.html
It's a one-liner to move the namespaces options into its own
sub-menu under 'General Setup'.
Oren.
On 10/07/2010 09:15 AM, Daniel Lezcano wrote:
The pid namespace
On 09/29/2010 09:17 PM, Matt Helsley wrote:
[ Persistent readers may wish to grab some coffee because this is gonna be
a long reply. ]
[ Persistent readers have been already warned ... ]
On Wed, Sep 29, 2010 at 06:22:37PM -0400, Oren Laadan wrote:
On 09/23/2010 05:53 PM, Matt Helsley
On 09/27/2010 06:03 PM, Brad Boyer wrote:
On Mon, Sep 27, 2010 at 12:16:28PM -0700, Matt Helsley wrote:
On Sun, Sep 26, 2010 at 12:08:37PM -0700, Brad Boyer wrote:
On Thu, Sep 23, 2010 at 02:53:28PM -0700, Matt Helsley wrote:
Not all filesystems will necessarily be able to support relinking
1 - 100 of 1787 matches
Mail list logo