cgroup notifications API and memory thresholds
cover cgroup notifications API and memory thresholds
It's my first attempt to implement cgroup notifications API and memory thresholds on top of it. The idea of API was proposed by Paul Menage.
It lacks some important features and need more testing, but I want publish it as soon as possible to get feedback from community.
TODO:
- memory thresholds on root cgroup;
- memsw support;
- documentation.
Kirill A. Shutemov (3):
cgroup: implement eventfd-based generic API for notifications
res_counter: implement thresholds
memcg: implement memory thresholds
include/linux/cgroup.h | 8 ++ include/linux/res_counter.h | 44 +++++++++++ kernel/cgroup.c | 181 ++++++++++++++++++++++++++++++++++++++++++- kernel/res_counter.c | 4 + mm/memcontrol.c | 149 +++++++++++++++++++++++++++++++++++ 5 files changed, 385 insertions(+), 1 deletions(-)
RFC PATCH 1/3 cgroup: implement eventfd-based generic API for notifications
This patch introduces write-only file "cgroup.event_control" in every cgroup.
To register new notification handler you need:
- create an eventfd;
- open a control file to be monitored. Callbacks register_event() and unregister_event() must be defined for the control file;
- write "<event_fd> <control_fd> <args>" to cgroup.event_control. Interpretation of args is defined by control file implementation;
eventfd will be woken up by control file implementation or when the cgroup is removed.
To unregister notification handler just close eventfd.
Signature · 1 line
1.6.5.3
A message that is not in the archive (sent off-list or not yet received): cover.1259248846.git.kirill-oKw7cIdHH8eLwutG50LtGA@public.gmane.org
RFC PATCH 1/3 cgroup: implement eventfd-based generic API for notifications
This patch introduces write-only file "cgroup.event_control" in every cgroup.
To register new notification handler you need:
- create an eventfd;
- open a control file to be monitored. Callbacks register_event() and unregister_event() must be defined for the control file;
- write "<event_fd> <control_fd> <args>" to cgroup.event_control. Interpretation of args is defined by control file implementation;
eventfd will be woken up by control file implementation or when the cgroup is removed.
To unregister notification handler just close eventfd.
Signature · 1 line
1.6.5.3
A message that is not in the archive (sent off-list or not yet received): bc4dc055a7307c8667da85a4d4d9d5d189af27d5.1259248846.git.kirill-oKw7cIdHH8eLwutG50LtGA@public.gmane.org
RFC PATCH 2/3 res_counter: implement thresholds
It allows to setup two thresholds: one above current usage and one below. Callback threshold_notifier() will be called if a threshold is crossed.
Signature · 1 line
1.6.5.3
A message that is not in the archive (sent off-list or not yet received): 8524ba285f6dd59cda939c28da523f344cdab3da.1259248846.git.kirill-oKw7cIdHH8eLwutG50LtGA@public.gmane.org
RFC PATCH 3/3 memcg: implement memory thresholds
It allows to register multiple memory thresholds and gets notifications when it crosses.
To register a threshold application need:
- create an eventfd;
- open file memory.usage_in_bytes of a cgroup
- write string "<event_fd> <memory.usage_in_bytes> <threshold>" to cgroup.event_control.
Application will be notified through eventfd when memory usage crosses threshold in any direction.
Signature · 1 line
1.6.5.3
Quoted 11 lines · Kirill A. Shutemov
It's my first attempt to implement cgroup notifications API and memory thresholds on top of it. The idea of API was proposed by Paul Menage.
It lacks some important features and need more testing, but I want publish it as soon as possible to get feedback from community.
TODO:
- memory thresholds on root cgroup;
- memsw support;
- documentation.
Maybe it would be interesting to do that for the /cgroup/<name>/tasks by sending in the event the number of tasks in the cgroup when it changes, so it more easy to detect 0 process event and then remove the cgroup directory, no ?
RFC PATCH 2/3 res_counter: implement thresholds
In reply to Kirill A. ShutemovIt allows to setup two thresholds: one above current usage and one below. Callback threshold_notifier() will be called if a threshold is crossed.
Signature · 1 line
1.6.5.3
RFC PATCH 3/3 memcg: implement memory thresholds
It allows to register multiple memory thresholds and gets notifications when it crosses.
To register a threshold application need:
- create an eventfd;
- open file memory.usage_in_bytes of a cgroup
- write string "<event_fd> <memory.usage_in_bytes> <threshold>" to cgroup.event_control.
Application will be notified through eventfd when memory usage crosses threshold in any direction.
Signature · 7 lines
1.6.5.3 -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to [email protected]. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: <a href=mailto:"[email protected]"> [email protected] </a>
A message that is not in the archive (sent off-list or not yet received): d977350fcc9bc3e1fe484440c1fc3a7470a4e26b.1259248846.git.kirill-oKw7cIdHH8eLwutG50LtGA@public.gmane.org
Re: [PATCH RFC v0 3/3] memcg: implement memory thresholds
Quoted 9 lines · Kirill A. Shutemov
It allows to register multiple memory thresholds and gets notifications when it crosses.
To register a threshold application need:
- create an eventfd;
- open file memory.usage_in_bytes of a cgroup
- write string "<event_fd> <memory.usage_in_bytes> <threshold>" to cgroup.event_control.
Quoted 5 lines
Application will be notified through eventfd when memory usage crosses threshold in any direction.
Signed-off-by Kirill A. Shutemov <[email protected]>
I don't see the patches attached or inlined in the emails that follow
Balbir
Re: [PATCH RFC v0 3/3] memcg: implement memory thresholds
In reply to Kirill A. ShutemovQuoted 9 lines · Kirill A. Shutemov
It allows to register multiple memory thresholds and gets notifications when it crosses.
To register a threshold application need:
- create an eventfd;
- open file memory.usage_in_bytes of a cgroup
- write string "<event_fd> <memory.usage_in_bytes> <threshold>" to cgroup.event_control.
Quoted 5 lines
Application will be notified through eventfd when memory usage crosses threshold in any direction.
Signed-off-by Kirill A. Shutemov <[email protected]>
I don't see the patches attached or inlined in the emails that follow
Balbir
A message that is not in the archive (sent off-list or not yet received): 661de9470911260903h4996070emec678f09a7ba2c9e-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org
Re: [PATCH RFC v0 3/3] memcg: implement memory thresholds
Quoted 11 lines · Balbir Singh
Quoted 9 lines · Kirill A. Shutemov
It allows to register multiple memory thresholds and gets notifications when it crosses.
To register a threshold application need:
- create an eventfd;
- open file memory.usage_in_bytes of a cgroup
- write string "<event_fd> <memory.usage_in_bytes> <threshold>" to cgroup.event_control.
Quoted 7 lines
Quoted 5 lines
Application will be notified through eventfd when memory usage crosses threshold in any direction.
Signed-off-by Kirill A. Shutemov <[email protected]>I don't see the patches attached or inlined in the emails that follow
Sorry. Resent.
_______________________________________________
Containers mailing list
[email protected]
https://lists.linux-foundation.org/mailman/listinfo/containers
Re: [PATCH RFC v0 3/3] memcg: implement memory thresholds
In reply to Balbir SinghQuoted 11 lines · Balbir Singh
Quoted 9 lines · Kirill A. Shutemov
It allows to register multiple memory thresholds and gets notifications when it crosses.
To register a threshold application need:
- create an eventfd;
- open file memory.usage_in_bytes of a cgroup
- write string "<event_fd> <memory.usage_in_bytes> <threshold>" to cgroup.event_control.
Quoted 7 lines
Quoted 5 lines
Application will be notified through eventfd when memory usage crosses threshold in any direction.
Signed-off-by Kirill A. Shutemov <[email protected]>I don't see the patches attached or inlined in the emails that follow
Sorry. Resent.
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to [email protected]. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"[email protected]"> [email protected] </a>
Quoted 11 lines · Kirill A. Shutemov
It's my first attempt to implement cgroup notifications API and memory thresholds on top of it. The idea of API was proposed by Paul Menage.
It lacks some important features and need more testing, but I want publish it as soon as possible to get feedback from community.
TODO:
- memory thresholds on root cgroup;
- memsw support;
- documentation.
Maybe it would be interesting to do that for the /cgroup/<name>/tasks by sending in the event the number of tasks in the cgroup when it changes, so it more easy to detect 0 process event and then remove the cgroup directory, no ?
Quoted 14 lines · Daniel Lezcano
Quoted 4 lines · Kirill A. Shutemov
It's my first attempt to implement cgroup notifications API and memory thresholds on top of it. The idea of API was proposed by Paul Menage.
Quoted 8 lines
It lacks some important features and need more testing, but I want publish it as soon as possible to get feedback from community.
TODO:
- memory thresholds on root cgroup;
- memsw support;
- documentation.
Quoted 4 lines
Maybe it would be interesting to do that for the /cgroup/<name>/tasks by sending in the event the number of tasks in the cgroup when it changes, so it more easy to detect 0 process event and then remove the cgroup directory, no ?
I'll do it later.
A message that is not in the archive (sent off-list or not yet received): [email protected]
Quoted 14 lines · Daniel Lezcano
Quoted 4 lines · Kirill A. Shutemov
It's my first attempt to implement cgroup notifications API and memory thresholds on top of it. The idea of API was proposed by Paul Menage.
Quoted 8 lines
It lacks some important features and need more testing, but I want publish it as soon as possible to get feedback from community.
TODO:
- memory thresholds on root cgroup;
- memsw support;
- documentation.
Quoted 4 lines
Maybe it would be interesting to do that for the /cgroup/<name>/tasks by sending in the event the number of tasks in the cgroup when it changes, so it more easy to detect 0 process event and then remove the cgroup directory, no ?
I'll do it later.
_______________________________________________
Containers mailing list
[email protected]
https://lists.linux-foundation.org/mailman/listinfo/containers