Added information about PID namespace to tracepoints
A message that is not in the archive (sent off-list or not yet received): [email protected]
Hello everyone,
We changed the patch code a bit to include all of the VTIDs of a task nested deep down in an arbitrary level of pid namespace. For example, if a task is running inside a container nested in another container, our patch will display the TID of the task from the point of view of :
1. The root container (or the host)
2. The parent container
3. The container the task is running into
We also rebased on master, and everything is running smooth and fine on our test machine (linux 3.19.3). It should support down to linux 3.8.1.
Here is an example of a field:
sched_process_fork: { cpu_id = 0 }, { parent_comm = "bash", parent_tid = 10739, parent_pid = 10739, parent_ns_inum = 4026532365, child_comm = "bash", child_tid = 10968, _vtids_length = 3, vtids = [ [0] = 10968, [1] = 2658, [2] = 1265 ], child_pid = 10968, child_ns_inum = 4026532365 }
Any comments are greatly appreciated!
Sebastien & Francis.
Quoted 30 lines · Sébastien Lorrain
Hello fellow LTTNG devs,
We are students from Polytechnique Montreal and we are currently working on an TraceCompass analysis module for Linux containers (LXC/Docker/Etc...). The information we track is mostly CPU usage by pid namespace, which would allow the identification of CPU-utilization related bottlenecks on a Linux container host.
We tried to come up with targeted information to recreate the container/PID namespace tree of a Linux host and we have modified lttng-modules to be able to do so :
https://github.com/Selora/lttng-modulesIn our analysis, we try to re-use as many information that was already available in the LTTNG kernel tracer. We build our container/namespace tree using the tasks and their parent recursively (using only the pid/vppid/ppid). However, we were unable to have a reliable model without some light modifications on some tracepoints. _/
The modifications to the lttng-modulues where the following:/_We added the PID namespace INode (from
/proc/PID/ns/pid) to the LTTNG statedump tracepoint.Also, to track new task/containers that would spawn during the tracing session, we also added multiple fields to the /sched_process_fork/ event :
* Added a VTID field for the children task. This is mandatory in our analysis, as we keep track of VTID/TID association.
* Added a/parent_ns_inum/ and /child_ns_inum/ field wich represent the pid namespaces inodes of the parent and child task respectively.
o The /parent_ns_inum/ is "not mandatory" in our analysis, but it keeps things simple as we don't have to track TID from parent containers and it keeps the code relatively independant wheter statedump is enabled or not.
o The /child_ns_inum///IS mandatory, because even if we keep track of the PID/VPID/PPID/VPPID that have spawned, it is possible to "inject" a task in an already existing namespace without repareting it to the child reaper of the container (this means the task is sent in a namespace, but it is not part of the process tree of the container of that namespace)
Quoted 9 lines
We hope to integrate our analysis to TraceCompass soon, and without the modification to the LTTNG tracer approved, we would be unable to proceed throught code review. We would be really grateful to the community if we could have feedback, and we will make every modifications possibles to have our analysis up and working!
The code is supposed to work on kernel version 3.8 through 3.19. It was tested on 3.18 and 3.19, and I'am going to test it for 3.8 today.
Cheers,
Sebastien & Francis.
From 46e5eb49eb6f7688d696ef7c7c5108a121c8af33 Mon Sep 17 00:00:00 2001
From: Lrouge <[email protected]>
Date: Wed, 15 Apr 2015 11:53:53 -0400
Subject: [PATCH] Added namespace info in sched_fork and statedump
Modifications to sched_fork :
Added a field for parent and child pid namespace inode (as in /proc/$PID/ns/pid). Also added a field for the child VTID.
Modifications to lttng_statedump_process_state :
Added a field for the pid namespace inode of the task.
--- .../events/lttng-module/lttng-statedump.h | 11 ++++ instrumentation/events/lttng-module/sched.h | 66 +++++++++++++++++++++- 2 files changed, 76 insertions(+), 1 deletion(-)
instrumentation/events/lttng-module/lttng-statedump.h +11 -0
@@ -8,6 +8,14 @@#include <linux/nsproxy.h> #include <linux/pid_namespace.h> #include <linux/types.h> +#include <linux/version.h> + + +#if (LINUX_VERSION_CODE >=KERNEL_VERSION (3 ,19 ,0 )) +#define lttng_proc_inum ns.inum +#else +#define lttng_proc_inum proc_inum +#endif LTTNG_TRACEPOINT_EVENT(lttng_statedump_start, TP_PROTO(struct lttng_session *session),
@@ -60,6 +68,9 @@ LTTNG_TRACEPOINT_EVENT(lttng_statedump_process_state, ctf_integer(int, submode, submode) ctf_integer(int, status, status) ctf_integer(int, ns_level, pid_ns ? pid_ns->level : 0) +#if (LINUX_VERSION_CODE >=KERNEL_VERSION (3 ,8 ,0 )) + ctf_integer(unsigned int, ns_inum, pid_ns ? pid_ns->lttng_proc_inum: 0) +#endif ) )
instrumentation/events/lttng-module/sched.h +65 -1
96 lines not on this page · raw message
Signature · 1 line
2.3.5
_______________________________________________
lttng-dev mailing list
[email protected]
http://lists.lttng.org/cgi-bin/mailman/listinfo/lttng-dev
A message that is not in the archive (sent off-list or not yet received): [email protected]
Patch merged with a few coding style fixes.
Thanks!
Mathieu
----- Original Message -----
Quoted 103 lines
Hello everyone,
We changed the patch code a bit to include all of the VTIDs of a task nested deep down in an arbitrary level of pid namespace.
For example, if a task is running inside a container nested in another container, our patch will display the TID of the task from the point of view of :1. The root container (or the host)
2. The parent container
3. The container the task is running intoWe also rebased on master, and everything is running smooth and fine on our test machine (linux 3.19.3).
It should support down to linux 3.8.1.Here is an example of a field :
sched_process_fork: {cpu_id= 0 }, {parent_comm= "bash",parent_tid= 10739,parent_pid= 10739,parent_ns_inum= 4026532365,child_comm= "bash",child_tid= 10968,_vtids_length= 3, vtids = [ [0] = 10968, [1] = 2658, [2] = 1265 ],child_pid= 10968,child_ns_inum= 4026532365 }Any comments are greatly appreciated!
Sebastien & Francis.
Quoted 73 lines · Sébastien Lorrain
Hello fellow LTTNG devs,
We are students from Polytechnique Montreal and we are currently working on an TraceCompass analysis module for Linux containers (LXC/Docker/Etc...). The information we track is mostly CPU usage by pid namespace, which would allow the identification of CPU-utilization related bottlenecks on a Linux container host.
We tried to come up with targeted information to recreate the container/PID namespace tree of a Linux host and we have modified lttng-modules to be able
to do so :https://github.com/Selora/lttng-modules
In our analysis, we try to re-use as many information that was already available in the LTTNG kernel tracer. We build our container/namespace tree using the tasks and their parent recursively (using only the pid/vppid/ppid). However, we were unable to have a reliable model without some light modifications on some tracepoints.
The modifications to the lttng-modulues where the following:
We added the PID namespace INode (from
/proc/PID/ns/pid) to the LTTNG statedump tracepoint.Also, to track new task/containers that would spawn during the tracing session, we also added multiple fields to the
sched_process_forkevent :* Added a VTID field for the children task. This is mandatory in our analysis, as we keep track of VTID/TID association.
* Added a
parent_ns_inumandchild_ns_inumfield wich represent the pid namespaces inodes of the parent and child task respectively.* The
parent_ns_inumis "not mandatory" in our analysis, but it keeps things
simple as we don't have to track TID from parent containers and it keeps the
code relatively independant wheter statedump is enabled or not.* The
child_ns_inumIS mandatory, because even if we keep track of the PID/VPID/PPID/VPPID that have spawned, it is possible to "inject" a task in an already existing namespace without repareting it to the child reaper of the container (this means the task is sent in a namespace, but it is not part of the process tree of the container of that namespace)We hope to integrate our analysis to TraceCompass soon, and without the modification to the LTTNG tracer approved, we would be unable to proceed throught code review.
We would be really grateful to the community if we could have feedback, and we will make every modifications possibles to have our analysis up and working!
The code is supposed to work on kernel version 3.8 through 3.19.
It was tested on 3.18 and 3.19, and I'am going to test it for 3.8 today.
Cheers,
Sebastien & Francis.
Signature · 3 lines
Mathieu Desnoyers EfficiOS Inc. http://www.efficios.com
_______________________________________________
lttng-dev mailing list
[email protected]
http://lists.lttng.org/cgi-bin/mailman/listinfo/lttng-dev