History log of /6.6.0/sigar/ (Results 1 - 25 of 2999)
Revision (<<< Hide revision tags) (Show revision tags >>>)Date Author Comments
(<<< Hide modified files)
(Show modified files >>>)
aa76848106-May-2020 Trond Norbye <trond.norbye@gmail.com>

MB-39149: Remove support for sigar_rpc_*

These methods cause inclusion of libtirpc which cause inclusion
of incompatible libraries on RHEL8.2

Change-Id: If1bd0af7915a9398de9b294

MB-39149: Remove support for sigar_rpc_*

These methods cause inclusion of libtirpc which cause inclusion
of incompatible libraries on RHEL8.2

Change-Id: If1bd0af7915a9398de9b2940c57741587431d630
Reviewed-on: http://review.couchbase.org/c/sigar/+/127445
Tested-by: Chris Hillery <ceej@couchbase.com>
Reviewed-by: Chris Hillery <ceej@couchbase.com>
Well-Formed: Build Bot <build@couchbase.com>

show more ...

e1b339d419-Mar-2020 Steve Watanabe <steve.watanabe@couchbase.com>

MB-38296 Report /proc/vmstat's allocstall stat

Sigar will return the following information from /proc/vmstat on linux
systems.

up until 4.10
allocstall,

MB-38296 Report /proc/vmstat's allocstall stat

Sigar will return the following information from /proc/vmstat on linux
systems.

up until 4.10
allocstall, /* up until 4.10 */

4.10 onwards the sum of the following
allocstall_dma, /* 4.10 onwards */
allocstall_dma32, /* 4.10 onwards */
allocstall_normal, /* 4.10 onwards */
allocstall_movable; /* 4.10 onwards */

If the stat isn't available and on non-linux systems it will return -1.

allocstall is the number of times direct reclaim occurs.

From https://lwn.net/Articles/396561/

Normally, memory pages are reclaimed (made available for new uses, with
data written back, if necessary) in the background; when all goes well,
there will always be a list of free pages available when memory is needed If,
however, a memory allocation request cannot be satisfied from the free list,
the kernel may try to reclaim pages directly in the context of the process
performing the allocation. Diverting an allocation request into this kind of
cleanup activity is called "direct reclaim."

This change requires the following ns_server change:
http://review.couchbase.org/#/c/124033/

Change-Id: I8a5be6d5a372e64f457b9cff69ac47a64cf2619e
Reviewed-on: http://review.couchbase.org/123991
Tested-by: Build Bot <build@couchbase.com>
Reviewed-by: Aliaksey Artamonau <aliaksey.artamonau@couchbase.com>

show more ...

b3ad9dcc26-Feb-2020 Aliaksey Artamonau <aliaksiej.artamonau@gmail.com>

Don't update cached process cpu info timestamp.

In cases when we just return the cached info. That way next time this
cached info is used to calculate cpu utilization, it's calculated ov

Don't update cached process cpu info timestamp.

In cases when we just return the cached info. That way next time this
cached info is used to calculate cpu utilization, it's calculated over
a correct interval (since when it was actually gathered, not since
when it was returned to the caller).

Change-Id: Iae031cd97174ef163e68ff758394ae62750a0228
Reviewed-on: http://review.couchbase.org/122824
Tested-by: Aliaksey Artamonau <aliaksey.artamonau@couchbase.com>
Tested-by: Build Bot <build@couchbase.com>
Reviewed-by: Steve Watanabe <steve.watanabe@couchbase.com>
Reviewed-by: Trond Norbye <trond.norbye@couchbase.com>

show more ...

6bd69c6425-Feb-2020 Aliaksey Artamonau <aliaksiej.artamonau@gmail.com>

Fix process cpu info staleness threshold.

If the comment around this code to be trusted, it was meant to be 1
second. In reality, it was 1 millisecond.

We do occasionally call s

Fix process cpu info staleness threshold.

If the comment around this code to be trusted, it was meant to be 1
second. In reality, it was 1 millisecond.

We do occasionally call sigar_proc_cpu_get on the same process in
quick succession. With the bad threshold, if less than 1 millisecond
passed between the calls, we'd get a cached garbage value for cpu
utilization (fixed by the preceding patch). Otherwise, we'd likely get
a zero (I can see that execution time stays the same in these
situations, but I'm not sure if it's the kernel that doesn't update
the information frequently enough, or if it's sigar caching the
information somewhere). This patch bumps the threshold to the intended
value of 1 second. Even with the fix the behavior is not perfect, but
it more or less works with how we use sigar.

Change-Id: Iaf782789f0e0a92a2a8d8ca74c239bfb58abb916
Reviewed-on: http://review.couchbase.org/122823
Tested-by: Aliaksey Artamonau <aliaksey.artamonau@couchbase.com>
Tested-by: Build Bot <build@couchbase.com>
Reviewed-by: Steve Watanabe <steve.watanabe@couchbase.com>
Reviewed-by: Trond Norbye <trond.norbye@couchbase.com>

show more ...

fdc80ea925-Feb-2020 Aliaksey Artamonau <aliaksiej.artamonau@gmail.com>

Maintain correct percent values in cached process cpu info.

In certain cases sigar_proc_cpu_get simply returns the cached
sigar_proc_cpu_t entry. The code as it was updated the cached

Maintain correct percent values in cached process cpu info.

In certain cases sigar_proc_cpu_get simply returns the cached
sigar_proc_cpu_t entry. The code as it was updated the cached
entry (memcpy(prev, proccpu, ...)) before setting proccpu->percent. So
the cached entry would have a garbage value for percent.

Change-Id: Ice3d9df0fd3fbf93d067a6ea1e521b0c9dcad752
Reviewed-on: http://review.couchbase.org/122822
Tested-by: Aliaksey Artamonau <aliaksey.artamonau@couchbase.com>
Reviewed-by: Steve Watanabe <steve.watanabe@couchbase.com>
Reviewed-by: Trond Norbye <trond.norbye@couchbase.com>
Tested-by: Build Bot <build@couchbase.com>

show more ...

e111352925-Feb-2020 Aliaksey Artamonau <aliaksiej.artamonau@gmail.com>

Don't leave gaps in replies when skipping processes.

The old code would always write the process information at the same
index as the process in the list of interesting processes. But th

Don't leave gaps in replies when skipping processes.

The old code would always write the process information at the same
index as the process in the list of interesting processes. But the
loop could skip some of the processes (for instance, when a process
has died). That would leave gaps in the reply. So a separate
index for the replies needs to be maintained.

Change-Id: I565627f2adad8bc5266c48b1be1ba06d88a9a26d
Reviewed-on: http://review.couchbase.org/122821
Tested-by: Aliaksey Artamonau <aliaksey.artamonau@couchbase.com>
Reviewed-by: Steve Watanabe <steve.watanabe@couchbase.com>
Reviewed-by: Trond Norbye <trond.norbye@couchbase.com>
Tested-by: Build Bot <build@couchbase.com>

show more ...

134a99de30-Jan-2020 Dave Rigby <daver@couchbase.com>

Fix warnings under Apple clang 11

Fix the following warnings when building under Apple clang version
11.0.0 (clang-1100.0.33.17) (as shipped with XCode 11.3.1):

../sigar/src

Fix warnings under Apple clang 11

Fix the following warnings when building under Apple clang version
11.0.0 (clang-1100.0.33.17) (as shipped with XCode 11.3.1):

../sigar/src/sigar_util.c:424:5: warning: extension used [-Wlanguage-extension-token]
asm volatile ("cpuid\n\t"
^
../sigar/src/os/darwin/darwin_sigar.c:246:32: warning: assigning to 'proc_pidinfo_func_t' (aka 'int (*)(int, int, unsigned long long, void *, int)') from 'void *' converts between void pointer and function pointer [-Wpedantic]
(*sigar)->proc_pidinfo = dlsym((*sigar)->libproc, "proc_pidinfo");
^ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
../sigar/src/os/darwin/darwin_sigar.c:247:34: warning: assigning to 'proc_pidfdinfo_func_t' (aka 'int (*)(int, int, int, void *, int)') from 'void *' converts between void pointer and function pointer [-Wpedantic]
(*sigar)->proc_pidfdinfo = dlsym((*sigar)->libproc, "proc_pidfdinfo");
^ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Change-Id: I769860b6e413a8ff97a45cf2bc1cd164d13efe9e
Reviewed-on: http://review.couchbase.org/121413
Tested-by: Build Bot <build@couchbase.com>
Reviewed-by: Jim Walker <jim@couchbase.com>

show more ...

c33791d615-Oct-2019 Aliaksey Artamonau <aliaksiej.artamonau@gmail.com>

Report more details about where CPU is spent.

Change-Id: I04220b70d07acdb772e5f77256531eec24a5ca1c
Reviewed-on: http://review.couchbase.org/116448
Tested-by: Build Bot <build@couchba

Report more details about where CPU is spent.

Change-Id: I04220b70d07acdb772e5f77256531eec24a5ca1c
Reviewed-on: http://review.couchbase.org/116448
Tested-by: Build Bot <build@couchbase.com>
Reviewed-by: Trond Norbye <trond.norbye@couchbase.com>
Tested-by: Aliaksey Artamonau <aliaksey.artamonau@couchbase.com>

show more ...

3671082814-Oct-2019 Aliaksey Artamonau <aliaksiej.artamonau@gmail.com>

Drop swap page in and page out stats.

Since ns_server doesn't actually use them.

Change-Id: I29d48a7e7c394c8960e8ea446e858ec6f81f4d07
Reviewed-on: http://review.couchbase.org/11

Drop swap page in and page out stats.

Since ns_server doesn't actually use them.

Change-Id: I29d48a7e7c394c8960e8ea446e858ec6f81f4d07
Reviewed-on: http://review.couchbase.org/116447
Tested-by: Build Bot <build@couchbase.com>
Reviewed-by: Trond Norbye <trond.norbye@couchbase.com>
Tested-by: Aliaksey Artamonau <aliaksey.artamonau@couchbase.com>

show more ...

dfce9bf419-Sep-2019 Trond Norbye <trond.norbye@gmail.com>

MB-34865: Just fetch the child processes of the given pid

To work around access problems on Mac (and simplify
the code) find all of the processes which have the
provided pid as one o

MB-34865: Just fetch the child processes of the given pid

To work around access problems on Mac (and simplify
the code) find all of the processes which have the
provided pid as one of its parents, and only try
to fetch more information from those pids.

Change-Id: I8a71b0222bc261afe3c122e0256d84784390a28e
Reviewed-on: http://review.couchbase.org/115039
Tested-by: Build Bot <build@couchbase.com>
Reviewed-by: Dave Rigby <daver@couchbase.com>

show more ...

eeb8b58428-Aug-2019 Trond Norbye <trond.norbye@gmail.com>

MB-34865: let sigar_os_proc_list_get only return my processes

sigar_port wants information about the child processes of the
babysitter, but the API to sigar won't let it just query just

MB-34865: let sigar_os_proc_list_get only return my processes

sigar_port wants information about the child processes of the
babysitter, but the API to sigar won't let it just query just
that. Instead it needs to walk over all processes and look at
the parent processes (but in order to fetch the ppid it collects
a lot of other stuff).

In sigar_os_proc_list_get we've got both the ppid and the uid
available, but to avoid changing too much in the code I figured
we could skip all process _not_ belonging to ourselves and then
hopefully get rid of any access restrictions later on.

Change-Id: I0d38ce5b91c471cac51e951a50e4062870c2ae30
Reviewed-on: http://review.couchbase.org/113948
Reviewed-by: Dave Rigby <daver@couchbase.com>
Tested-by: Build Bot <build@couchbase.com>
Reviewed-by: Daniel Owen <owend@couchbase.com>

show more ...

7bf5fce329-Aug-2019 Trond Norbye <trond.norbye@gmail.com>

readdir_r is deprecated, use readdir instead

This is stated in the man page for readdir_r on linux

Change-Id: Ia84f1ce68e4701eff424ebf5be98d19443b8bfcd
Reviewed-on: http://revie

readdir_r is deprecated, use readdir instead

This is stated in the man page for readdir_r on linux

Change-Id: Ia84f1ce68e4701eff424ebf5be98d19443b8bfcd
Reviewed-on: http://review.couchbase.org/114009
Reviewed-by: Dave Rigby <daver@couchbase.com>
Tested-by: Build Bot <build@couchbase.com>

show more ...

3d3397e624-Jul-2019 Aliaksey Artamonau <aliaksiej.artamonau@gmail.com>

Take babysitter pid as a command line parameter.

Erlang 20 starts all processes via an intermediate process called
erl_child_setup. This breaks the old logic of finding out the
babys

Take babysitter pid as a command line parameter.

Erlang 20 starts all processes via an intermediate process called
erl_child_setup. This breaks the old logic of finding out the
babysitter pid by going two levels up in the process hierarchy. And as
a consequence of this, we don't collect information about processes
like memcached.

This change addresses this by accepting the babysitter pid on the
command line instead of trying to deduce it.

Change-Id: Ia1e103f5a40ff6205a8feaf9ea85fcefe7689300
Reviewed-on: http://review.couchbase.org/112378
Tested-by: Build Bot <build@couchbase.com>
Reviewed-by: Dave Rigby <daver@couchbase.com>
Reviewed-by: Artem Stemkovski <artem@couchbase.com>
Tested-by: Aliaksey Artamonau <aliaksey.artamonau@couchbase.com>

show more ...

57d360a530-Nov-2018 Dave Rigby <daver@couchbase.com>

MB-32205: Fix heap-buffer-overflow in find_interesting_procs

When searching for 'interesting' processes, sigar looks for child
processes matching a given parent process, and iterates acr

MB-32205: Fix heap-buffer-overflow in find_interesting_procs

When searching for 'interesting' processes, sigar looks for child
processes matching a given parent process, and iterates across all
child processes. However there is a potential heap overflow during the
iteration - it can dereference past the end of the array (procs_end)
via child->ppid, as it checks for reaching the end of the array
*after* dereferencing the pointer.

Fix by flipping the order of the check.

Change-Id: I0e146407a06c0d8df06403a26a0d875b9468e8df
Reviewed-on: http://review.couchbase.org/102328
Reviewed-by: Trond Norbye <trond.norbye@gmail.com>
Reviewed-by: Sergey Avseyev <sergey.avseyev@gmail.com>
Tested-by: Build Bot <build@couchbase.com>

show more ...

73353fe613-Mar-2018 Sergey Avseyev <sergey.avseyev@gmail.com>

Use libtirpc on Fedora Linux 28+

https://fedoraproject.org/wiki/Changes/SunRPCRemoval

Change-Id: Iacb9dca35b48d0e6080cf173035e4956c62ec3ea
Reviewed-on: http://review.couchbase.o

Use libtirpc on Fedora Linux 28+

https://fedoraproject.org/wiki/Changes/SunRPCRemoval

Change-Id: Iacb9dca35b48d0e6080cf173035e4956c62ec3ea
Reviewed-on: http://review.couchbase.org/90845
Tested-by: Sergey Avseyev <sergey.avseyev@gmail.com>
Well-Formed: Build Bot <build@couchbase.com>
Tested-by: Build Bot <build@couchbase.com>
Reviewed-by: Dave Rigby <daver@couchbase.com>

show more ...

79ee063516-Aug-2016 Sergey Avseyev <sergey.avseyev@gmail.com>

Backport patch for modern FreeBSD

https://github.com/freebsd/freebsd-ports/commit/98d85d7795618efb7df88471f78f7ff0d70a85d4

Change-Id: Ic104aca6ccebdb555c57480b77b9f7330379df9e
R

Backport patch for modern FreeBSD

https://github.com/freebsd/freebsd-ports/commit/98d85d7795618efb7df88471f78f7ff0d70a85d4

Change-Id: Ic104aca6ccebdb555c57480b77b9f7330379df9e
Reviewed-on: http://review.couchbase.org/66792
Tested-by: Sergey Avseyev <sergey.avseyev@gmail.com>
Reviewed-by: Trond Norbye <trond.norbye@gmail.com>
Tested-by: buildbot <build@couchbase.com>

show more ...

83bd885928-Aug-2015 Trond Norbye <trond.norbye@gmail.com>

Refactor: remove unused function tcp_curr_estab

Change-Id: I998a97773780cf3ca3a74d70b3fdcafd16a7b1bc
Reviewed-on: http://review.couchbase.org/54896
Tested-by: buildbot <build@couchba

Refactor: remove unused function tcp_curr_estab

Change-Id: I998a97773780cf3ca3a74d70b3fdcafd16a7b1bc
Reviewed-on: http://review.couchbase.org/54896
Tested-by: buildbot <build@couchbase.com>
Reviewed-by: Dave Rigby <daver@couchbase.com>

show more ...

3c5e5f6328-Aug-2015 Trond Norbye <trond.norbye@gmail.com>

Refactor: remove unused "getpass" code

Change-Id: Iacb18b251a4a2d5b5756eb02fb2410b88a3e0b46
Reviewed-on: http://review.couchbase.org/54895
Tested-by: buildbot <build@couchbase.com>

Refactor: remove unused "getpass" code

Change-Id: Iacb18b251a4a2d5b5756eb02fb2410b88a3e0b46
Reviewed-on: http://review.couchbase.org/54895
Tested-by: buildbot <build@couchbase.com>
Reviewed-by: Dave Rigby <daver@couchbase.com>

show more ...

b7c3b4ad28-Aug-2015 Trond Norbye <trond.norbye@gmail.com>

Refactor: remove unused function sys_info_get_uname

Change-Id: I2533ab5ed95654569084a0bb3918021deb0cb178
Reviewed-on: http://review.couchbase.org/54894
Tested-by: buildbot <build@cou

Refactor: remove unused function sys_info_get_uname

Change-Id: I2533ab5ed95654569084a0bb3918021deb0cb178
Reviewed-on: http://review.couchbase.org/54894
Tested-by: buildbot <build@couchbase.com>
Reviewed-by: Dave Rigby <daver@couchbase.com>

show more ...

7038e53c27-Aug-2015 Trond Norbye <trond.norbye@gmail.com>

MB-16166: Use libproc to get mem_resident on macosx

Change-Id: Ide37c10771bc1656f5e703ff3273a47d146ef73c
Reviewed-on: http://review.couchbase.org/54890
Tested-by: buildbot <build@cou

MB-16166: Use libproc to get mem_resident on macosx

Change-Id: Ide37c10771bc1656f5e703ff3273a47d146ef73c
Reviewed-on: http://review.couchbase.org/54890
Tested-by: buildbot <build@couchbase.com>
Reviewed-by: Dave Rigby <daver@couchbase.com>

show more ...

Revision tags: v4.0.0
829a4c0313-May-2015 Trond Norbye <trond.norbye@gmail.com>

Remove tons of unused code

Change-Id: I1ebf8e9538d8aa2ed62967a81d46d083d05f78ad
Reviewed-on: http://review.couchbase.org/50974
Tested-by: buildbot <build@couchbase.com>
Reviewed-

Remove tons of unused code

Change-Id: I1ebf8e9538d8aa2ed62967a81d46d083d05f78ad
Reviewed-on: http://review.couchbase.org/50974
Tested-by: buildbot <build@couchbase.com>
Reviewed-by: Dave Rigby <daver@couchbase.com>

show more ...

3b99ab9113-May-2015 Trond Norbye <trond.norbye@gmail.com>

No need to set binary dir

Change-Id: I2c46eedb2d7c2a42a60f357a07e21b46fb0f6ba6
Reviewed-on: http://review.couchbase.org/50973
Tested-by: buildbot <build@couchbase.com>
Reviewed-b

No need to set binary dir

Change-Id: I2c46eedb2d7c2a42a60f357a07e21b46fb0f6ba6
Reviewed-on: http://review.couchbase.org/50973
Tested-by: buildbot <build@couchbase.com>
Reviewed-by: Dave Rigby <daver@couchbase.com>

show more ...

76dad86d19-Feb-2015 Aliaksej Artamonau <aliaksiej.artamonau@gmail.com>

Pass process ppid with process info.

Change-Id: Ibeed81a24cafef99e0dd271573c487df90f3377a
Reviewed-on: http://review.couchbase.org/47055
Tested-by: Aliaksey Artamonau <aliaksiej.arta

Pass process ppid with process info.

Change-Id: Ibeed81a24cafef99e0dd271573c487df90f3377a
Reviewed-on: http://review.couchbase.org/47055
Tested-by: Aliaksey Artamonau <aliaksiej.artamonau@gmail.com>
Reviewed-by: Aliaksey Kandratsenka <alkondratenko@gmail.com>

show more ...

332532a219-Feb-2015 Aliaksej Artamonau <aliaksiej.artamonau@gmail.com>

Increase process name length.

Change-Id: I259bbee5361b549470f45a8173e11a3a97f686d1
Reviewed-on: http://review.couchbase.org/47054
Tested-by: Aliaksey Artamonau <aliaksiej.artamonau@g

Increase process name length.

Change-Id: I259bbee5361b549470f45a8173e11a3a97f686d1
Reviewed-on: http://review.couchbase.org/47054
Tested-by: Aliaksey Artamonau <aliaksiej.artamonau@gmail.com>
Reviewed-by: Aliaksey Kandratsenka <alkondratenko@gmail.com>

show more ...

42d0b98417-Feb-2015 Trond Norbye <trond.norbye@gmail.com>

Fix up whitespace issues

Change-Id: I8f365161f6b093aaf0deab3db49fef8f078bde45
Reviewed-on: http://review.couchbase.org/46965
Tested-by: buildbot <build@couchbase.com>
Reviewed-by

Fix up whitespace issues

Change-Id: I8f365161f6b093aaf0deab3db49fef8f078bde45
Reviewed-on: http://review.couchbase.org/46965
Tested-by: buildbot <build@couchbase.com>
Reviewed-by: Dave Rigby <daver@couchbase.com>

show more ...

12345678910>>...120