Activity
From 04/07/2015 to 05/06/2015
05/06/2015
- 05:59 PM Bug #2251: Spread multicast communication not captured
- IPv4 at least working now. What is lacking is the detection whether we have a multicast message for IPv6. Should be d...
- 03:29 PM Bug #2251 (Rejected): Spread multicast communication not captured
- Due to the fact that the receiving address of a multicast package is not the address of a specific computer but inste...
- 05:58 PM Revision 3221e4ef (process-monitor): Handle IPv4 Multicast
- There needs to be special handling for multicast because the destination
address of a multicast package is not the ad... - 02:01 PM Revision a4384f75 (process-monitor): Allow compilation with 0.11
- 11:17 AM Revision 7649cac1 (process-monitor): Do not convert boost filesystem exceptions to invalid_argument
- These ones are not caught
05/05/2015
- 04:41 PM Revision 5a524389 (process-monitor): Allow boost filesystem exceptions from socket parsing
- These are caught anyway instead of invalid_argument exceptions
04/21/2015
- 06:40 PM Revision 81da9af5 (db-adapter): Add command line usage to the documentation
- 02:28 PM Revision 3c1d65b3 (db-adapter): Swap aggregation and derivative calculation
- This make the derivative calculation easier and prevents potential
calculation problems in case of newly created subp... - 12:09 PM Revision 22061dd1 (db-adapter): Specify rst dependencies correctly
04/20/2015
- 06:33 PM Revision ae9924bc (db-adapter): Add a basic sphinx documentation
- 06:31 PM Revision 0a0ab203 (db-adapter): Missing ignores
- 05:51 PM Revision 5d1bf380 (db-adapter): Rename to rsbperfmon-db-adapter
- This finally reflects the actual thing done by this code.
- 05:38 PM Revision 789af41f (db-adapter): Code cleanup
- * Pylint warnings
* Commenting using Google style docstrings
04/16/2015
04/15/2015
- 06:19 PM Bug #2189 (Rejected): key error in influxdb adapter
- This was probably caused by one of my experiments and shouldn't happen during normal operation.
- 06:18 PM Bug #2221 (Resolved): NetBandwidthProvider does not notice that a process terminated
- Resolved via merge
- 06:11 PM Revision e5459125 (process-monitor): Reorganize files and folders for better separation
- 05:17 PM Revision 0d80f984 (process-monitor): Port to boost signals2
- 04:42 PM Revision f8f07c78 (process-monitor): Merge branch 'wip-network-rewrite'
- (cherry picked from commit f2ab10e043c488eb60ed5664bd3647815ba46852)
- 04:42 PM Revision f2ab10e0 (process-monitor): Merge branch 'wip-network-rewrite'
- 04:30 PM Revision aca3c0bf (process-monitor): Handle termination correctly
- Prevent segfaults and deadlocks from threading issues.
- 02:06 PM Revision 112c9235 (process-monitor): Notice dying processes
- refs #2221
- 01:15 PM Revision e7a8c17c (process-monitor): Use pcacp blocking mode
- We have a separate thread for dispatching which is allowed to block.
- 12:17 PM Revision a53f075d (process-monitor): Performance optimization
- Replace exceptions with null pointer checks for a considerable speedup
- 11:59 AM Revision 8ecf095f (process-monitor): Bug hunting
- * join ipv4 and ipv6 processing because of missed connection due to
compatibility modes
* Fix caching strategy and ...
04/14/2015
- 02:57 PM Revision e47a940c (process-monitor): Implement caching for inode -> process mapping
- Considerable speed up
- 12:15 PM Revision 320543ef (process-monitor): Interface for process lookup
- Prepare for a caching implementation.
- 11:36 AM Revision bf0fa103 (process-monitor): Optimize comparison operators for connections
- Since we are using maps frequently for connection lookup, operators need
to be fast.
04/13/2015
- 09:07 PM Revision 210bb75a (process-monitor): Implement connection caching
- Adds basic caching for open connection to prevent constant proc file
system access and parsing. - 08:07 PM Revision 88c842a3 (process-monitor): Interface for network connection pools
- Provide an object oriented interface to acquire and manage open network
connections so that different implementations... - 07:36 PM Revision c620b23a (process-monitor): Replace nethogs code with custom implementation
- Account for loopback device and associate traffic correctly to received
and sent counters.
This is a very inefficien...
04/09/2015
- 01:28 PM Bug #2221 (Resolved): NetBandwidthProvider does not notice that a process terminated
- This would result in a lot of useless processing instead of stopping the thread that monitored the process.
04/08/2015
- 04:57 PM Revision b2b594a7 (db-adapter): Make the influxdb time series prefixes configurable
- (cherry picked from commit e8ac647d5e9c5598c8c52239838860bdb2a5935d)
- 04:48 PM Revision e8ac647d (db-adapter): Make the influxdb time series prefixes configurable
- 04:03 PM Revision fa7beb61 (process-monitor): Send monitored pid:name pairs via RSB
- This allows to control the remote mode from process-based instances.
(cherry picked from commit 3a9bbd3797f7b7bfb8bd... - 04:03 PM Revision e114d99b (process-monitor): Implement a remote-controlled process monitor
- Adds a new operational mode which allows to send rsb messages to control
the PIDs to monitor instead of using the com... - 03:56 PM Revision 3a9bbd37 (process-monitor): Send monitored pid:name pairs via RSB
- This allows to control the remote mode from process-based instances.
- 03:00 PM Revision 19e87519 (process-monitor): Implement a remote-controlled process monitor
- Adds a new operational mode which allows to send rsb messages to control
the PIDs to monitor instead of using the com... - 02:33 PM Revision 3fae69e1 (process-monitor): Use a more reasonable logger name for process monitor
Also available in: Atom