statsd histogram to support more than 1 second
Summary:
Problem: Collectd/Statsd supports configurable percentiles for timers but it limits the value to be 1 to 1000ms. If the timer value is more than 1000ms, it uses for min, max, average,... and drops it for percentile computation.
Solution: Added support for increasing bin width when the value is above histogram's range.
Test Plan: Tested by sending metrics within range and out-of-range (ie above 1000 ms)
Reviewers: skhajamo, shalstea
Reviewed By: skhajamo
CC: arcyd
Differential Revision: https://all.phab.dev.bloomberg.com/D156454
Summary:
Problem: Collectd/Statsd supports configurable percentiles for timers but it limits the value to be 1 to 1000ms. If the timer value is more than 1000ms, it uses for min, max, average,... and drops it for percentile computation.
Solution: Added support for increasing bin width when the value is above histogram's range.
Test Plan: Tested by sending metrics within range and out-of-range (ie above 1000 ms)
Reviewers: skhajamo, shalstea
Reviewed By: skhajamo
CC: arcyd
Differential Revision: https://all.phab.dev.bloomberg.com/D156454
Merge branch 'collectd-5.3' into collectd-5.4
Conflicts:
src/write_graphite.c
Conflicts:
src/write_graphite.c
Merge branch 'collectd-4.10' into collectd-5.3
Revert "Let snmp_synch_response deal with PDU freeing"
This reverts commit 9d6f72e5054cfbb183585991fefcf119259d71c4.
The proper fix to this issue is in the collectd-4.10 branch. See also
79e90bba.
This reverts commit 9d6f72e5054cfbb183585991fefcf119259d71c4.
The proper fix to this issue is in the collectd-4.10 branch. See also
79e90bba.
Revert "Let snmp_synch_response deal with PDU freeing"
This reverts commit e767dc103a142e1c57800f06a9c384f21a48aef5.
The proper fix to this issue is in the collectd-4.10 branch. See also
79e90bba.
This reverts commit e767dc103a142e1c57800f06a9c384f21a48aef5.
The proper fix to this issue is in the collectd-4.10 branch. See also
79e90bba.
snmp: avoid freeing req under normal operation
Equivalent patch to 79e90bba5, to avoid issue #804 introduced while
fixing #610.
Equivalent patch to 79e90bba5, to avoid issue #804 introduced while
fixing #610.
Let snmp_synch_response deal with PDU freeing
When reading from tables, upon errors the PDUs sent are already
freed by snmp_synch_response since they are right after
snmp_send is called.
This commit syncs collectd's approach with other occurences of
snmp_synch_response calls.
There might be a few corner cases where we leak PDUs, but it
is unclear how to check for those since we would need to
have an indication that snmp_send was never called, which
as far as I can tell is not possible.
The potential for failure in snmp_send is rather low and will
be easily spotted though, since when crafting invalid PDUs
snmp send will constantly fail and since valid configurations
can never leak memory.
This fixes #804
When reading from tables, upon errors the PDUs sent are already
freed by snmp_synch_response since they are right after
snmp_send is called.
This commit syncs collectd's approach with other occurences of
snmp_synch_response calls.
There might be a few corner cases where we leak PDUs, but it
is unclear how to check for those since we would need to
have an indication that snmp_send was never called, which
as far as I can tell is not possible.
The potential for failure in snmp_send is rather low and will
be easily spotted though, since when crafting invalid PDUs
snmp send will constantly fail and since valid configurations
can never leak memory.
This fixes #804
Let snmp_synch_response deal with PDU freeing
When reading from tables, upon errors the PDUs sent are already
freed by snmp_synch_response since they are right after
snmp_send is called.
This commit syncs collectd's approach with other occurences of
snmp_synch_response calls.
There might be a few corner cases where we leak PDUs, but it
is unclear how to check for those since we would need to
have an indication that snmp_send was never called, which
as far as I can tell is not possible.
The potential for failure in snmp_send is rather low and will
be easily spotted though, since when crafting invalid PDUs
snmp send will constantly fail and since valid configurations
can never leak memory.
This fixes #804
When reading from tables, upon errors the PDUs sent are already
freed by snmp_synch_response since they are right after
snmp_send is called.
This commit syncs collectd's approach with other occurences of
snmp_synch_response calls.
There might be a few corner cases where we leak PDUs, but it
is unclear how to check for those since we would need to
have an indication that snmp_send was never called, which
as far as I can tell is not possible.
The potential for failure in snmp_send is rather low and will
be easily spotted though, since when crafting invalid PDUs
snmp send will constantly fail and since valid configurations
can never leak memory.
This fixes #804
Merge pull request #796 from mfournier/write_redis-colon-mistake
write_redis: re-add colon dropped in b7984797 + manpage
write_redis: re-add colon dropped in b7984797 + manpage
Link aggregation.so to libm.so
Conflicts:
src/Makefile.am
Conflicts:
src/Makefile.am
Merge pull request #765 from tokkee/sh/write-err
Use the complain mechanism to report write failures.
Use the complain mechanism to report write failures.
network: don't enable gcrypt thread callbacks when gcrypt recent enough
From `gcrypt.h`:
> NOTE: Since Libgcrypt 1.6 the thread callbacks are not anymore used.
> However we keep it to allow for some source code compatibility if used
> in the standard way.
Otherwise, we get a deprecation warning which is turned into an error:
```
CC libcollectdclient_la-network_buffer.lo
../../../src/libcollectdclient/network_buffer.c:58:15: error: 'gcry_thread_cbs' is deprecated (declared at /usr/include/gcrypt.h:213) [-Werror=deprecated-declarations]
GCRY_THREAD_OPTION_PTHREAD_IMPL;
```
Fixes: #632
From `gcrypt.h`:
> NOTE: Since Libgcrypt 1.6 the thread callbacks are not anymore used.
> However we keep it to allow for some source code compatibility if used
> in the standard way.
Otherwise, we get a deprecation warning which is turned into an error:
```
CC libcollectdclient_la-network_buffer.lo
../../../src/libcollectdclient/network_buffer.c:58:15: error: 'gcry_thread_cbs' is deprecated (declared at /usr/include/gcrypt.h:213) [-Werror=deprecated-declarations]
GCRY_THREAD_OPTION_PTHREAD_IMPL;
```
Fixes: #632
add write_redis to manpage
write_redis: re-add colon dropped in b7984797
When running f3706b0b87, the following command gets sent to redis:
"ZADD" "collectd/hostname/entropy/entropy" "1415487432.000000" "1415487432:932"
Meaning the value actually stored, and later returned by redis is:
"<timstamp>:<value>".
b7984797 accidentally dropped the comma separating the timestamp and the
value, which leads the plugin to store a somewhat confusing value in
redis:
"ZADD" "collectd/hostname/entropy/entropy" "1415487432.000000" "1415487432932"
When running f3706b0b87, the following command gets sent to redis:
"ZADD" "collectd/hostname/entropy/entropy" "1415487432.000000" "1415487432:932"
Meaning the value actually stored, and later returned by redis is:
"<timstamp>:<value>".
b7984797 accidentally dropped the comma separating the timestamp and the
value, which leads the plugin to store a somewhat confusing value in
redis:
"ZADD" "collectd/hostname/entropy/entropy" "1415487432.000000" "1415487432932"
redhat initscript: add timeout when stopping the daemon
This mitigates the risk of having 2 daemons running simultaneously
during a short period, on restart.
The debian package's initscript does the same thing, using the same
timeout value.
This mitigates the risk of having 2 daemons running simultaneously
during a short period, on restart.
The debian package's initscript does the same thing, using the same
timeout value.
Merge pull request #764 from tokkee/sh/collectd-5.3
write_graphite plugin: Improved error reporting when connecting to the server.
write_graphite plugin: Improved error reporting when connecting to the server.
fix grammar in plugin loading error message
"cause for this problem are" => "cause for this problem is"
"cause for this problem are" => "cause for this problem is"
logfile: avoid total silence in case of a misconfiguration
Basically the same patch that was applied to syslog.c in 84c38056
Basically the same patch that was applied to syslog.c in 84c38056
correct typo in error message
Syslog: if we can't find the loglevel specified by the configuration string default to 'info' and warn about the unknown configuration option. no way to make syslog totaly silent anymore.
curl_json: fix the array access implemented in f1e1e37e
In the avl-tree we store two different structs, cj_key_t for the value
we search, c_avl_tree_t for sub-nodes.
The old version does assume when it will find a key, and when a tree,
which doesn't have to be right in all cases.
Therefore we utilize the magic to revalidate this cast.
Being able to tell tree from key, we now can also implement array access
on the right most node of the tree-path.
In the avl-tree we store two different structs, cj_key_t for the value
we search, c_avl_tree_t for sub-nodes.
The old version does assume when it will find a key, and when a tree,
which doesn't have to be right in all cases.
Therefore we utilize the magic to revalidate this cast.
Being able to tell tree from key, we now can also implement array access
on the right most node of the tree-path.
Use the complain mechanism to report write failures.
This ensures that we don't spam the log for each submitted value. Instead, an
exponential backup is used to report permanent errors starting at one message
per second.
This ensures that we don't spam the log for each submitted value. Instead, an
exponential backup is used to report permanent errors starting at one message
per second.
write_graphite plugin: Improved error reporting when connecting to the server.
statsd: clarify/simplify documentation & example
Added documentation for StatsD module per added functionality in issue #403.
Bit shift right by 8 to get return code from exit status
Pull-Request: #741
Signed-off-by: Florian Forster <octo@collectd.org>
Pull-Request: #741
Signed-off-by: Florian Forster <octo@collectd.org>
cpu plugin: Remove temperature code for Mac OS X.
This code would throw an error for some systems, because the temperature
couldn't be read. Since the plugin still works in general, this was very
confusing to users. Becaus the temperature, even when read sucessfully, was
never dispatched, remove the code entirely. We might want to add this back to
the apple_sensors plugin eventually.
Fixes: #22
This code would throw an error for some systems, because the temperature
couldn't be read. Since the plugin still works in general, this was very
confusing to users. Becaus the temperature, even when read sucessfully, was
never dispatched, remove the code entirely. We might want to add this back to
the apple_sensors plugin eventually.
Fixes: #22
swap plugin: Make "cached" optional under Linux.
This is the smallest possible fix for OpenVZ, where cached is not available.
master has a more complete fix which should be used going forward.
Fixes: #733
This is the smallest possible fix for OpenVZ, where cached is not available.
master has a more complete fix which should be used going forward.
Fixes: #733
Fix typo in if caluse
Fixes #729
Signed-off-by: Florian Forster <octo@collectd.org>
Fixes #729
Signed-off-by: Florian Forster <octo@collectd.org>
Merge branch 'collectd-5.3' into collectd-5.4
src/common.c: Fix strstripnewline().
Because the length of the string was decremented too late, the function
would actually overwrite the trailing null byte instead of the newline.
Because the length of the string was decremented too late, the function
would actually overwrite the trailing null byte instead of the newline.
RPM specfile: enable lvm plugin on RHEL7
Merge branch 'collectd-5.3' into collectd-5.4
Conflicts:
contrib/redhat/collectd.spec
Conflicts:
contrib/redhat/collectd.spec
RPM specfile: enable netlink plugin on RHEL > 5
RPM specfile: fix bogus dates
modern rpmbuild versions appear to be more strict on this point.
modern rpmbuild versions appear to be more strict on this point.
RPM specfile: support for RHEL7
RPM specfile: remove define accidentally re-added when handling merge conflict
RPM specfile: remove macro accidentally re-added when handling merge conflict
RPM specfile: remove duplicate modbus block
Merge branch 'collectd-5.3' into collectd-5.4
Conflicts:
src/curl.c
Conflicts:
src/curl.c
network plugin: Improve client connecting behavior.
This moves the socket creation logic so it's called from
networt_send_buffer_plain(). This allows us to recover after network
failures or when collectd was started before the network was available.
Fixes: #627
This moves the socket creation logic so it's called from
networt_send_buffer_plain(). This allows us to recover after network
failures or when collectd was started before the network was available.
Fixes: #627
src/utils_match.[ch]: Implement match_value_reset().
This is called from the curl and memcachec plugins. Previously, gauge
average, minimum and maximum values were calculated since the start of
the daemon, rather than per-interval. This is not the documented
functionality.
Fixes: #663
This is called from the curl and memcachec plugins. Previously, gauge
average, minimum and maximum values were calculated since the start of
the daemon, rather than per-interval. This is not the documented
functionality.
Fixes: #663
curl plugin: Use the utility functions for config handling.
Merge remote-tracking branch 'github/collectd-5.3' into collectd-5.3
src/meta_data.c: If the meta_data_t is empty, avoid calling malloc(0).
src/utils_format_json.c: Handle empty meta_data_t gracefully.
Previously, `meta != NULL` and `keys_num == 0` was possible and would
result in "{}", which is not valid JSON.
Fixes: 716
Previously, `meta != NULL` and `keys_num == 0` was possible and would
result in "{}", which is not valid JSON.
Fixes: 716
python plugin: Don't create empty meta_data_t objects.
Fixes: #716
Fixes: #716
src/utils_rrdcreate.c: Free "args".
Thanks to Yves Mettier for pointing this out!
Fixes: #661
Thanks to Yves Mettier for pointing this out!
Fixes: #661
Merge branch 'collectd-5.3' into collectd-5.4
Conflicts:
contrib/redhat/collectd.spec
Conflicts:
contrib/redhat/collectd.spec
RPM specfile: update changelog without incrementing release number
GenericJMXConfConnection: Remove unused imports.
java plugin: Make sure cjni_thread_detach() is called on all paths.
Each call to cjni_thread_attach() much be accompanied by
a call to cjni_thread_detach(). Some error handling cases were missing
the call, potentially locking the plugin up.
Each call to cjni_thread_attach() much be accompanied by
a call to cjni_thread_detach(). Some error handling cases were missing
the call, potentially locking the plugin up.
collectd.conf(5): Improve AutoLoadPlugin documentation.
Make it clear that plugins without a <Plugin /> block will still need a
LoadPlugin line.
References: #715
Make it clear that plugins without a <Plugin /> block will still need a
LoadPlugin line.
References: #715
collectd.conf(5): Improve documentation of LoadPlugin.
Mention AutoLoadPlugin in the relevant places. Add a <Plugin /> block
for the "df" plugin to make it easier to spot that it is preceded by a
<LoadPlugin /> *block*.
Fixes: #715
Mention AutoLoadPlugin in the relevant places. Add a <Plugin /> block
for the "df" plugin to make it easier to spot that it is preceded by a
<LoadPlugin /> *block*.
Fixes: #715
EL <6 provide perl-ExtUtils-Embed with base perl
Conflicts:
contrib/redhat/collectd.spec
Conflicts:
contrib/redhat/collectd.spec
Statsd: support samplerate field in timing metric types
Fixes "Unable to parse line" bug, often seen receiving stats from
OpenStack Swift. Should fix GH issue #461
Fixes "Unable to parse line" bug, often seen receiving stats from
OpenStack Swift. Should fix GH issue #461
collectd.conf: Fix default packet size of the network plugin.
Fixes: #710
Fixes: #710
Merge branch 'collectd-5.3' into collectd-5.4
Conflicts:
contrib/redhat/collectd.spec
Conflicts:
contrib/redhat/collectd.spec
Merge branch 'collectd-4.10' into collectd-5.3
curl_xml plugin: Fixed tautological pointer comparison error.
Fixes: http://bugs.debian.org/758481
Signed-off-by: Florian Forster <octo@collectd.org>
Fixes: http://bugs.debian.org/758481
Signed-off-by: Florian Forster <octo@collectd.org>
collectd.conf(5): add detail & example about multi-instance filterchain targets
snmp: free snmp_pdu struct allocated by snmp_pdu_create()
This should fix the leak reported in issue #610.
This should fix the leak reported in issue #610.
Merge pull request #540 from tokkee/configure
configure: Fixed check for libsigrok and AM_PATH_GLIB_2_0.
configure: Fixed check for libsigrok and AM_PATH_GLIB_2_0.
Merge pull request #689 from ymettier/ym/cd53_fix_tcpconns_inet_diag
Fix for 'linux/inet_diag.h: No such file or directory'
Fix for 'linux/inet_diag.h: No such file or directory'
RedHat RPM spec: add modbus plugin
... now that EPEL6 provides libmodbus
... now that EPEL6 provides libmodbus
Fix for 'linux/inet_diag.h: No such file or directory'
fix: use DOUBLE_TO_CDTIME_T() to match input type
... as strtod() returns a double not a time_t.
Follow-up to 7d1d59fb to match what ada80db25 does (currently in
master) and avoid a merge conflict.
... as strtod() returns a double not a time_t.
Follow-up to 7d1d59fb to match what ada80db25 does (currently in
master) and avoid a merge conflict.
mistake, SAVE_LIBS="$LIBS" is correct.
patch required to build oracle plugin, collectd-5.4.0 (#653)
Merge pull request #579 from mfournier/libiptc-ldflags
link iptables plugin against flags from pkg-config
link iptables plugin against flags from pkg-config
Merge pull request #677 from collectd/revert-630-F-METRICS-383-tsdb-writer-5.4
Revert "write_tsdb plugin"
Revert "write_tsdb plugin"
Revert "write_tsdb plugin"
Merge pull request #630 from llnw/F-METRICS-383-tsdb-writer-5.4
write_tsdb plugin
write_tsdb plugin
utils_vl_lookup: Fixed a race when creating user objects.
This could cause multiple aggregation instances to be created in the
aggregation plugin when first writing data to the plugin. This, in turn, led
to "value too old" warnings because subsequently all data was submitted twice.
Thanks to @faxm0dem for reporting this in GH #535.
This could cause multiple aggregation instances to be created in the
aggregation plugin when first writing data to the plugin. This, in turn, led
to "value too old" warnings because subsequently all data was submitted twice.
Thanks to @faxm0dem for reporting this in GH #535.
Plugin exec: change notification time to integer.
Acording to collectd-exec.5 NOTIFICATION DATA FORMAT: time is epoch, i. e. as seconds since 1970-01-01 00:00:00 UTC.
Acording to collectd-exec.5 NOTIFICATION DATA FORMAT: time is epoch, i. e. as seconds since 1970-01-01 00:00:00 UTC.
cmd PUTNOTIF time option: handle double values.
Conflicts:
src/utils_cmd_putnotif.c
Conflicts:
src/utils_cmd_putnotif.c
Plugin java: fix notification time conversion to cdtime_t.
Plugin threshold: time is not set in missing notifications.
Merge pull request #592 from mfournier/bug-582
curl_json: also allow long callback names.
curl_json: also allow long callback names.
manpage: document 2 missing options to amqp plugin section
add missing backticks which broke the build
write_riemann: coding style consistency
Make the style in line with 50d1b34d6 (from master) and avoid later
merge conflict.
Make the style in line with 50d1b34d6 (from master) and avoid later
merge conflict.
don't assume pkg-config is in PATH
Configparser: when we alocate an empty list, we also need to reset the counter; else we will trip over this later.
This fixes https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=750440
This fixes https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=750440
[METRICS-390] Make sure HostTags is initialized if not configured
[METRICS-390] Fix string handling for TSDB prefix and tags
[METRICS-390] Add tag writing from metadata using TSDB target_set patches
[METRICS-383] Add HostTags which adds tags for all metrics from this writer
Uncrustify write_tsdb
Add DeleteSocket option in synopsis of manpage
Signed-off-by: Florian Forster <octo@collectd.org>
Signed-off-by: Florian Forster <octo@collectd.org>
[METRICS-383] write_tsdb plugin
add brackets to the list of characters graphite chokes on
Graphite deals poorly with metric names containing quotes.
Sanitize graphite metrics with our own escaping function,
better suited for this use-case than the one in `utils_parse_option`
Sanitize graphite metrics with our own escaping function,
better suited for this use-case than the one in `utils_parse_option`
curl_json: also allow long callback names.
This is basically the same as 2f392819, which got applied to the
curl_xml plugin a few months ago.
Fixes #582.
This is basically the same as 2f392819, which got applied to the
curl_xml plugin a few months ago.
Fixes #582.
Fix check for htonll in configure.ac
fix for #477: Notification time gone wrong
Sub-second time handling was added in collectd 5.0.0. This adds support
for this to the target_notification plugin, which got missed out at the
time.
Sub-second time handling was added in collectd 5.0.0. This adds support
for this to the target_notification plugin, which got missed out at the
time.
Fix getargs declaration to compile in AIX7
AIX doesn't have MSG_DONTWAIT, define as MSG_NONBLOCK
Updated Color to match DataSources and DSName for single-series graphs.