On Tue, Feb 09, 2016 at 07:06:01PM +0300, Cyrill Gorcunov wrote:
...
> I'll go this way and patch criu stable.
Here is a patch for stable, just to not loose it.
>From a7a6c115f1e2aa95f7cae727b465427746078b63 Mon Sep 17 00:00:00 2001
From: Cyrill Gorcunov
Date: Tue, 9 Feb 2016 22:15:55 +0300
Subje
On Tue, Feb 09, 2016 at 06:58:59PM +0300, Igor Sukhih wrote:
>
> this is c/r duty to handle time natively.
Yes
> It will be solved somehow in the future in mainstream.
Yes. Via custom cgroups engine, we simply have to add
description into criu command line.
> but now it resolved in Va7 by 've'
On 02/09/2016 06:51 PM, Cyrill Gorcunov wrote:
On Tue, Feb 09, 2016 at 06:46:07PM +0300, Igor Sukhih wrote:
Guys if we agree on this, I'll send updated variant for libvzctl.
it not needed to update libvzctl.
it should be custom vz specific patch to criu to handle time specific
cgroups natively
On Tue, Feb 09, 2016 at 06:46:07PM +0300, Igor Sukhih wrote:
> >Guys if we agree on this, I'll send updated variant for libvzctl.
>
> it not needed to update libvzctl.
> it should be custom vz specific patch to criu to handle time specific
> cgroups natively
I already explained, criu should not be
On 02/09/2016 04:56 PM, Cyrill Gorcunov wrote:
On Tue, Feb 09, 2016 at 04:43:53PM +0300, Cyrill Gorcunov wrote:
This two members represent monotonic and bootbased clocks for
container's uptime. When container is in suspended state (or
moving to another node) we trest monotonic and bootbased
cloc
Vladimir Davydov writes:
> Currently, we only throttle on reading/writing page cache. As a result,
> if a user runs e.g. `find` inside a container, it will quickly consume
> all available quota, so that a following attempt to read/write a file
> will hang for an unacceptably long period of time i
On 09.02.2016 17:19, Dmitry Monakhov wrote:
> - If (count % BITS_PER_PAGE == 0) we stuck in endless loop.
> - (!set) case was also wrong
>
> https://jira.sw.ru/browse/PSBM-43936
>
> Signed-off-by: Dmitry Monakhov
Reviewed-by:Vasily Averin
> ---
> block/blk-cbt.c |7 +++
> 1 files
- If (count % BITS_PER_PAGE == 0) we stuck in endless loop.
- (!set) case was also wrong
https://jira.sw.ru/browse/PSBM-43936
Signed-off-by: Dmitry Monakhov
---
block/blk-cbt.c |7 +++
1 files changed, 3 insertions(+), 4 deletions(-)
diff --git a/block/blk-cbt.c b/block/blk-cbt.c
index
On Tue, Feb 09, 2016 at 04:43:53PM +0300, Cyrill Gorcunov wrote:
> This two members represent monotonic and bootbased clocks for
> container's uptime. When container is in suspended state (or
> moving to another node) we trest monotonic and bootbased
> clocks as being stopped so we need to account
This two members represent monotonic and bootbased clocks for
container's uptime. When container is in suspended state (or
moving to another node) we trest monotonic and bootbased
clocks as being stopped so we need to account delta time
on restore and adjust the members in subject.
Moreover this t
On Tue, Feb 09, 2016 at 03:53:00PM +0300, Pavel Emelyanov wrote:
> > file names into CRIU options so c/r will be done
> > automatically.
>
> For now we can keep a patch in vz-criu spec file for that, can't we?
I already do that. I've sent patches to Igor for libvzctl.
Now I propose to rename the
On 12/04/2015 08:16 PM, Vasily Averin wrote:
> Customer reports that free or top command inside Centos7 CT shows 0 memory
> used
>
> free from procps-ng-3.3.10-3.el7 uses /proc/meminfo output and calculates
> used = total - free - buffers - cache - slab
>
> However kernel in some cases calcula
On 02/09/2016 02:25 PM, Cyrill Gorcunov wrote:
> On Tue, Feb 09, 2016 at 02:14:25PM +0300, Vladimir Davydov wrote:
>>>
>>> It was named so in pcs6, so I tried to preserve naming. But sure I
>>> can rename, that's not a problem.
>>
>> I mean not ve_struct.start_timespec - its name is perfectly fine
On Tue, Feb 09, 2016 at 02:14:25PM +0300, Vladimir Davydov wrote:
> >
> > It was named so in pcs6, so I tried to preserve naming. But sure I
> > can rename, that's not a problem.
>
> I mean not ve_struct.start_timespec - its name is perfectly fine - but
> the cgroup file ve.start_timespec - with
On Tue, Feb 09, 2016 at 01:53:04PM +0300, Cyrill Gorcunov wrote:
> On Tue, Feb 09, 2016 at 01:29:56PM +0300, Vladimir Davydov wrote:
> > >
> > > Yes. We take current values then substract the @real_[start_]timespec
> > > and return it. IOW they are just monotonic|boottime clocks.
> >
> > Then I t
On Tue, Feb 09, 2016 at 01:29:56PM +0300, Vladimir Davydov wrote:
> >
> > Yes. We take current values then substract the @real_[start_]timespec
> > and return it. IOW they are just monotonic|boottime clocks.
>
> Then I think we should name these files accordingly to avoid confusion.
> ve.clock_{m
On Tue, Feb 09, 2016 at 01:12:42PM +0300, Cyrill Gorcunov wrote:
> On Tue, Feb 09, 2016 at 01:02:09PM +0300, Vladimir Davydov wrote:
> > On Mon, Feb 08, 2016 at 11:32:12AM +0300, Cyrill Gorcunov wrote:
> > > +static int ve_ts_read(struct cgroup *cg, struct cftype *cft, struct
> > > seq_file *m)
>
On Tue, Feb 09, 2016 at 01:02:09PM +0300, Vladimir Davydov wrote:
> On Mon, Feb 08, 2016 at 11:32:12AM +0300, Cyrill Gorcunov wrote:
> > +static int ve_ts_read(struct cgroup *cg, struct cftype *cft, struct
> > seq_file *m)
> > +{
> > + struct ve_struct *ve = cgroup_ve(cg);
> > + struct timespe
On Mon, Feb 08, 2016 at 11:32:12AM +0300, Cyrill Gorcunov wrote:
> +static int ve_ts_read(struct cgroup *cg, struct cftype *cft, struct seq_file
> *m)
> +{
> + struct ve_struct *ve = cgroup_ve(cg);
> + struct timespec ts, *delta;
> +
> + do_posix_clock_monotonic_gettime(&ts);
> + i
Since they can be easily dropped at any moment, tcache pages should
contribute to MemAvailable.
Signed-off-by: Vladimir Davydov
---
fs/proc/meminfo.c | 8
mm/tcache.c | 9 +++--
2 files changed, 15 insertions(+), 2 deletions(-)
diff --git a/fs/proc/meminfo.c b/fs/proc/meminfo
Currently, we only throttle on reading/writing page cache. As a result,
if a user runs e.g. `find` inside a container, it will quickly consume
all available quota, so that a following attempt to read/write a file
will hang for an unacceptably long period of time in order to compensate
for the burst
Signed-off-by: Vladimir Davydov
---
include/bc/task.h | 12
1 file changed, 12 deletions(-)
diff --git a/include/bc/task.h b/include/bc/task.h
index bd723b86949e..e639a458776e 100644
--- a/include/bc/task.h
+++ b/include/bc/task.h
@@ -17,9 +17,7 @@ struct callback_head;
#ifdef CONF
22 matches
Mail list logo
