Performance improvment : uc_get_names now allocates memory by blocks and not element by element.
swap plugin: Add the "ReportBytes" option.
When enabled, swap I/O is reported in bytes, not pages. Only valid for Linux.
This should fix Github issue #10.
When enabled, swap I/O is reported in bytes, not pages. Only valid for Linux.
This should fix Github issue #10.
Merge branch 'collectd-5.1'
Merge branch 'collectd-5.0' into collectd-5.1
Conflicts:
src/swap.c
Conflicts:
src/swap.c
Merge branch 'collectd-4.10' into collectd-5.0
corrected high-resolution time conversions after code review
Signed-off-by: Florian Forster <octo@collectd.org>
Signed-off-by: Florian Forster <octo@collectd.org>
Updated Adrian Perez' 2009 Python network proto classes
Signed-off-by: Florian Forster <octo@collectd.org>
Signed-off-by: Florian Forster <octo@collectd.org>
snmp plugin: Fix error message printed by the "Shift" option.
I noticed a small error in snmp.c in collectd-5.1.0 (might be
elsewhere). The warning message in csnmp_config_add_data_shift says
"Scale config option" instead of "Shift" config option:
[…]
Signed-off-by: Florian Forster <octo@collectd.org>
I noticed a small error in snmp.c in collectd-5.1.0 (might be
elsewhere). The warning message in csnmp_config_add_data_shift says
"Scale config option" instead of "Shift" config option:
[…]
Signed-off-by: Florian Forster <octo@collectd.org>
df plugin: when checking / do it just once
Newer system symlink /etc/mtab to /proc/mounts which makes root
filesystem seem mounted twice, once as type rootfs and once for the
real device.
When / is selected for df plugin, skip the rootfs mount-point as
real root is usually mounted right on top of it.
Not skipping it causes rrdtool to complain about two updates for the
same timestamp, the first entry for rootfs passes but rrdtool complains
when stats for the second occurrence for real filesystem is being
considered.
Signed-off-by: Bruno Prémont <bonbons@linux-vserver.org>
Signed-off-by: Florian Forster <octo@collectd.org>
Newer system symlink /etc/mtab to /proc/mounts which makes root
filesystem seem mounted twice, once as type rootfs and once for the
real device.
When / is selected for df plugin, skip the rootfs mount-point as
real root is usually mounted right on top of it.
Not skipping it causes rrdtool to complain about two updates for the
same timestamp, the first entry for rootfs passes but rrdtool complains
when stats for the second occurrence for real filesystem is being
considered.
Signed-off-by: Bruno Prémont <bonbons@linux-vserver.org>
Signed-off-by: Florian Forster <octo@collectd.org>
snmp plugin: Check for SNMP_ENDOFMIBVIEW.
Some implementations of SNMP will send a special "variable" when you're
leaving the requested subtree with type "endOfMibView" aka.
"SNMP_ENDOFMIBVIEW". Examples of such implementations are Citrix
Netscalers and IBM's General Parallel File System (GPFS). Reading past
this special variable will result in weird values being returned.
This adds an appropriate check to csnmp_check_res_left_subtree().
Thanks to "Mark" for reporting this issue and clarifying.
Some implementations of SNMP will send a special "variable" when you're
leaving the requested subtree with type "endOfMibView" aka.
"SNMP_ENDOFMIBVIEW". Examples of such implementations are Citrix
Netscalers and IBM's General Parallel File System (GPFS). Reading past
this special variable will result in weird values being returned.
This adds an appropriate check to csnmp_check_res_left_subtree().
Thanks to "Mark" for reporting this issue and clarifying.
src/Makefile.am: Unconditionally link collectd with libm.
Under FreeBSD, running src/utils_format_json.c fails because isfinite()
cannot be resolved. The GNU manpages for isnan() and isfinite() also
advise users to link with libm and AIX has had the library being linked
in for a while. Since it's such a basic library it seems easier to just
link with it unconditionally.
Fixes Github issue #122.
Under FreeBSD, running src/utils_format_json.c fails because isfinite()
cannot be resolved. The GNU manpages for isnan() and isfinite() also
advise users to link with libm and AIX has had the library being linked
in for a while. Since it's such a basic library it seems easier to just
link with it unconditionally.
Fixes Github issue #122.
Merge pull request #124 from tokkee/bj/build
Fix some build issues on AIX using automake <= 1.9 and gcc 4.7.1.
Fix some build issues on AIX using automake <= 1.9 and gcc 4.7.1.
Add missing return statements
Fix gcc warn/error: no return statement in function returning non-void [-Werror=return-type]
Signed-off-by: Sebastian Harl <sh@tokkee.org>
Fix gcc warn/error: no return statement in function returning non-void [-Werror=return-type]
Signed-off-by: Sebastian Harl <sh@tokkee.org>
Automake 1.9 compatibility: don't use $(builddir)
Automake <= 1.9 don't define $(builddir) which is always defined to
'.' anyway.
Signed-off-by: Sebastian Harl <sh@tokkee.org>
Automake <= 1.9 don't define $(builddir) which is always defined to
'.' anyway.
Signed-off-by: Sebastian Harl <sh@tokkee.org>
src/collectd.conf.in: Update memcached config block.
memcached plugin: Fix some error messages.
memcached plugin: Fix backwards compatibility.
Don't use the instance name in the identifier when we're using the legacy
configuration mode.
Don't use the instance name in the identifier when we're using the legacy
configuration mode.
memcached plugin: Reorder functions to avoid prototype.
memcached plugin: Refactor the memcached_query_daemon() function.
The connecting code has been broken out in separate functions and the
writing and reading from the socket no longer uses poll(2),
non-blocking I/O and a custom built retry logic. Instead block on I/O and
let the read-thread-pool do its thing.
The connecting code has been broken out in separate functions and the
writing and reading from the socket no longer uses poll(2),
non-blocking I/O and a custom built retry logic. Instead block on I/O and
let the read-thread-pool do its thing.
memcached plugin: Fix backwards compatibility.
Register a default callback in the init() phase if no configuration is
provided for the plugin. This way doing just "LoadPlugin memcached" will
result in the old behavior i.e. localhost being queried.
Register a default callback in the init() phase if no configuration is
provided for the plugin. This way doing just "LoadPlugin memcached" will
result in the old behavior i.e. localhost being queried.
memcached plugin: Fix the default behavior, i.e. use the documented host and port.
collectd.conf(5): Document the memcached plugin.
The Instance blocks and Socket option were missing.
The Instance blocks and Socket option were missing.
memcached plugin: Specify read-group name when registering callbacks.
memcached plugin: Rename config callback (add prefix).
memcached plugin: Replace config_set_string() with cf_util_get_string().
multiple memcached instances, v2
memcached plugin : added support for multiple instances
Merge branch 'sh/collectd-5.1' into collectd-5.1
Merge branch 'ms/tcpconns'
tcpconns plugin: Only chose between netlink / proc on the first iteration.
Rather than going for netlink each iteration and fall back to /proc, try
this on the first iteration only and stick with the result afterwards.
Also contains some improvements to the error handling / reporting.
Rather than going for netlink each iteration and fall back to /proc, try
this on the first iteration only and stick with the result afterwards.
Also contains some improvements to the error handling / reporting.
Added the instance name to varnish API
Fixed issue collectd#110
Signed-off-by: Florian Forster <octo@collectd.org>
Fixed issue collectd#110
Signed-off-by: Florian Forster <octo@collectd.org>
little style changes (code review)
src/utils_format_json.c: Remove unused arguments.
write_graphite: Use the broken out "format_graphite" module.
This was broken out for use in the AMQP plugin. De-duplicate the code by
using it from write_graphite as well.
This was broken out for use in the AMQP plugin. De-duplicate the code by
using it from write_graphite as well.
AUTHORS: Add Thomas Meson.
Add Florian Forster to the list of copyright holders and authors of
utils_format_graphite.c, since its based on his code.
Add Florian Forster to the list of copyright holders and authors of
utils_format_graphite.c, since its based on his code.
amqp plugin: add support for Graphite output
This commit implements "Graphite format" for AMQP Plugin.
The AMQP plugin will be able to directly output a valid
Graphite metric format (<metric name> <value> <timestamp>\n).
This is very useful when the Graphite server is directly reading
from an AMQP broker. You can then avoid having a proxy somewhere
doing the conversion between PUTVAL or JSON metrics into Graphite
format.
Signed-off-by: Florian Forster <octo@collectd.org>
This commit implements "Graphite format" for AMQP Plugin.
The AMQP plugin will be able to directly output a valid
Graphite metric format (<metric name> <value> <timestamp>\n).
This is very useful when the Graphite server is directly reading
from an AMQP broker. You can then avoid having a proxy somewhere
doing the conversion between PUTVAL or JSON metrics into Graphite
format.
Signed-off-by: Florian Forster <octo@collectd.org>
Merge pull request #109 from tokkee/sh/collectd-5.0
configure: Define _REENTRANT when compiling on Solaris.
LGTM
configure: Define _REENTRANT when compiling on Solaris.
LGTM
Fix to make write_mongodb work with any version of mongo-c-driver
tcpconns/linux: Use netlink instead of parsing /proc/net/tcp{,6}
configure: Define _REENTRANT when compiling on Solaris.
On Wed, Mar 16, 2011 at 10:35:07AM +0100, Mathijs Möhlmann wrote:
> I'm using collectd 4.10.2 on Solaris 10 (gcc 3.4.6). Sometimes when I
> start with a clean rrd directory or add a host the .rrd file (rrdtools
> plugin) don't get created and I get the following message:
>
> collectd[2996]: [ID 702911 daemon.error] stat(/usr/local/var/lib/collectd/rrd/asterix/load/load.rrd) failed: Bad file number
This patch fixes this.
Signed-off-by: Sebastian Harl <sh@teamix.net>
On Wed, Mar 16, 2011 at 10:35:07AM +0100, Mathijs Möhlmann wrote:
> I'm using collectd 4.10.2 on Solaris 10 (gcc 3.4.6). Sometimes when I
> start with a clean rrd directory or add a host the .rrd file (rrdtools
> plugin) don't get created and I get the following message:
>
> collectd[2996]: [ID 702911 daemon.error] stat(/usr/local/var/lib/collectd/rrd/asterix/load/load.rrd) failed: Bad file number
This patch fixes this.
Signed-off-by: Sebastian Harl <sh@teamix.net>
configure, utils_dns: Check for and use netinet/ip_compat.h.
This is required on Solaris in order to get the ip6_ext struct.
This is required on Solaris in order to get the ip6_ext struct.
mysql plugin: Use the database instance name for identifying read callbacks.
The database instance is specified as argument to the <Database> block.
Without this fix, we cannot specify two database blocks without the
(optional!) 'Database' option as this would result in registering two read
callbacks with the name 'mysql'.
Thanks to Johan Bergström for reporting this on IRC!
The database instance is specified as argument to the <Database> block.
Without this fix, we cannot specify two database blocks without the
(optional!) 'Database' option as this would result in registering two read
callbacks with the name 'mysql'.
Thanks to Johan Bergström for reporting this on IRC!
contrib/migrate-4-5.px: Added missing end quotes in a call to rrd_filter.px.
Fix bad definition of UDP src port in DNS plugin.
Fix bad definition of UDP src port in DNS plugin.
collectd.conf(5): Document the "Password" option.
redis plugin: Coding style fixes.
redis plugin: Add authentication support.
Correct compilation on FreeBSD, set HOST_NAME_MAX to _POSIX_HOST_NAME_MAX if isn't defined
src/utils_format_json.c: Various fixes for the meta data code.
* Moved to a separate function.
* Escape string values.
* Fix format string for [u]int64_t.
* Emit the JSON literals «true» and «false» rather than «0» and «1».
* Make certain keys are always freed.
* Move the meta data information into a separate map.
* Moved to a separate function.
* Escape string values.
* Fix format string for [u]int64_t.
* Emit the JSON literals «true» and «false» rather than «0» and «1».
* Make certain keys are always freed.
* Move the meta data information into a separate map.
annotate meta data to json objects
Run through any metadata and annotate them to json objects when they are
created.
Run through any metadata and annotate them to json objects when they are
created.
contrib/migrate-4-5.px: Break up "df" RRD files into multiple files.
collectd.conf(5): Improve description of the configuration syntax.
Merge pull request #97 from tokkee/sh/swap
swap plugin: Don't report an error if there is no swap space on Linux.
swap plugin: Don't report an error if there is no swap space on Linux.
swap plugin: Don't report an error if there is no swap space on Linux.
… rather, simply record zeros in that case. Swap may be switched on and off at
arbitrary times and/or might be added at "later" times. Thus, storing zero in
case the swap plugin is enabled sounds like the best approach to me.
… rather, simply record zeros in that case. Swap may be switched on and off at
arbitrary times and/or might be added at "later" times. Thus, storing zero in
case the swap plugin is enabled sounds like the best approach to me.
Merge pull request #96 from mstap/master
add collectd systemd service file
add collectd systemd service file
describe collectd.service in contrib/README
Add collectd systemd unit file
Fix for config file parsing ("memcachec plugin: Option `Server' not allowed here.")
Fix out-of-bound index in rrdtool.c
Fix a reference to argv[1] that is (always?)
outside the bounds of the argv array. It happens
during a log statement that seems to want
to actually refer to the rrd filename.
Fix a reference to argv[1] that is (always?)
outside the bounds of the argv array. It happens
during a log statement that seems to want
to actually refer to the rrd filename.
oracle plugin: Add database name and query name (if applicable) to error messages.
oracle plugin: Report the "connect id" when OCILogon() fails.
collection3: Use {instance} in Df's grpah title.
{type_instance} didn't work when used with other plugins than "df", e.g.
the example configuration of the Oracle plugin uses this.
{type_instance} didn't work when used with other plugins than "df", e.g.
the example configuration of the Oracle plugin uses this.
oracle plugin: Replace the o_config_set_string() function.
oracle plugin: Implement the "Host" option.
src/utils_dns.c: Remove duplicate line.
src/utils_dns.c: Rename the "s_addr" and "d_addr" variables.
On Solaris, there is a macro called "s_addr" resulting in syntax errors
when this is used as a variable name. The variables have been renamed to
circumvent this problem. Thanks to Maciej Bliziński for reporting this
problem and helping to debug it!
On Solaris, there is a macro called "s_addr" resulting in syntax errors
when this is used as a variable name. The variables have been renamed to
circumvent this problem. Thanks to Maciej Bliziński for reporting this
problem and helping to debug it!
collection3: Filter out identifiers with control characters.
Otherwise XML errors will prevent the page from rendering.
Otherwise XML errors will prevent the page from rendering.
src/utils_avltree.c: Add sanity check to c_avl_destroy().
src/utils_rrdcreate.c: Handle "last update" in a saner way.
netapp plugin: Close connection on communication errors.
collectd.conf(5): Fix typo.
collectdctl: Fix PUTVAL for data sets with multiple data sources.
Thanks to Cyril Feraudet for pointing out this problem!
Thanks to Cyril Feraudet for pointing out this problem!
contrib/collection3: Ignore control characters.
While they are correctly encoded, e.g. as "", browsers will
complain about invalid XML and not render the page. Ignore all control
characters other than whitespace.
While they are correctly encoded, e.g. as "", browsers will
complain about invalid XML and not render the page. Ignore all control
characters other than whitespace.
Create directories with permission 0777.
Let umask work its magic. Thanks to "Poil" for reporting this problem.
Fixes GitHub issue #63.
Let umask work its magic. Thanks to "Poil" for reporting this problem.
Fixes GitHub issue #63.
varnish plugin: Make argument "stats" const.
ethstat plugin: Fix a typo.
Bump version to 5.1.0; Update ChangeLog.
Merge branch 'collectd-5.0'
Merge branch 'collectd-4.10' into collectd-5.0
Conflicts:
ChangeLog
Conflicts:
ChangeLog
ChangeLog: Correct date.
Bump version to 5.0.4; Update ChangeLog.
Merge branch 'collectd-4.10' into collectd-5.0
Conflicts:
ChangeLog
src/collectd.conf.pod
src/common.c
src/network.c
src/processes.c
version-gen.sh
Conflicts:
ChangeLog
src/collectd.conf.pod
src/common.c
src/network.c
src/processes.c
version-gen.sh
Bump version to 4.10.7; Update ChangeLog.
Revert "Proposal fix for large "MaxPacketSize" use."
This reverts commit 12213f8dfeee3feb3846aa658def0bb79c1211d0.
This reverts commit 12213f8dfeee3feb3846aa658def0bb79c1211d0.
network plugin: Do not forward received notifications.
The network plugin would send out all notifications it got through its
callback, even the ones it received itself and even when forwarding was
disabled.
With this patch the network plugin will only transmit notifications created
locally -- forwarding of notifications is not implemented due to missing
loop-detection. For loop-detection we would have to keep track of the last
timestamp of each notifications, and since each host/plugin/type/-instance
field may be empty, this is not as straight forward as with value lists. If
someone needs this, they will need to invest some work I'm afraid.
The network plugin would send out all notifications it got through its
callback, even the ones it received itself and even when forwarding was
disabled.
With this patch the network plugin will only transmit notifications created
locally -- forwarding of notifications is not implemented due to missing
loop-detection. For loop-detection we would have to keep track of the last
timestamp of each notifications, and since each host/plugin/type/-instance
field may be empty, this is not as straight forward as with value lists. If
someone needs this, they will need to invest some work I'm afraid.
Merge pull request #56 from octo/ci/nfs
NFS plugin: Add Solaris support.
NFS plugin: Add Solaris support.
Merge branch 'ff/zfs'
zfs_arc plugin: Move reading of kstat values into separate functions.
This way error handling can be done in a central place and we can do this more
carefully overall.
This way error handling can be done in a central place and we can do this more
carefully overall.
Merge pull request #60 from octo/ar/zfs
more zfs kstat values
more zfs kstat values
zfs_arc plugin: Add copyright header for Aurelien Rougemont.
Proposal fix for large "MaxPacketSize" use.
When we use a "MaxPacketSize" over two thousand of octets (to pass huge
message in notification for example), buffer may not be flushed for a
while.
By flushing buffer when there is about 1400 octets, we're sure there is
no data too longer in.
When we use a "MaxPacketSize" over two thousand of octets (to pass huge
message in notification for example), buffer may not be flushed for a
while.
By flushing buffer when there is about 1400 octets, we're sure there is
no data too longer in.
zfs_arc plugin: Fix the "hash collision" and "mutex miss" metrics.
zfs_arc plugin: Fix typo in variable name.
zfs_arc plugin: Introduce the single-valued "mutex_operation" type.
zfs_arc plugin: Introduce the single-valued "cache_eviction" type.
zfs_arc plugin: Convert to use DERIVE.
add types for deleted, evict, hash, mutex
Signed-off-by: Florian Forster <octo@collectd.org>
Signed-off-by: Florian Forster <octo@collectd.org>
add hash_collisions, deleted, mutex_miss, evict_*
Fixed Solaris compilation errors
The value returned by get_kstat_value is of type long long whereas values is a union. Fixed the code so that the value gets assigned to one of the union members, in this case counter.
The value returned by get_kstat_value is of type long long whereas values is a union. Fixed the code so that the value gets assigned to one of the union members, in this case counter.