eclipse-sailing-analytics/sailing-analytics
1
0 Fork 0
mirror of https://github.com/eclipse-sailing-analytics/sailing-analytics.git synced 2026-10-16 09:11:02 +00:00
Code Activity
8,403 Commits 274 Branches 255 Tags
4d53a3d865a52d6661da8a34bd01cc124d836fc3
Commit Graph
7084 Commits
Author SHA1 Message Date
Axel Uhl 4d53a3d865 fixed smart future cache test which now failed due to more rigid locking policy; 
can't suspend the cache(...) method which is holding the write lock while trying to read using get(...).
Now suspending the computeCacheUpdate method. This still lets get(...) read null, but during
computeCacheUpdate the write lock for the key isn't held by the cache yet. Therefore, the get can proceed.
 2013-03-29 15:43:29 +01:00
Axel Uhl 465d618fbc fixed smart future cache test which now failed due to more rigid locking policy; 
can't suspend the cache(...) method which is holding the write lock while trying to read using get(...).
The get(...) method will simply block because a cache update is currently on its way.
 2013-03-29 15:40:01 +01:00
Axel Uhl 12c9b3761b fixed third of the leaderboard tests that failed due to the changed incurred by the fix for bug 1023 2013-03-29 15:15:49 +01:00
Axel Uhl 8922ccf470 fixed second leaderboard test case that broke because of the changes incurred by the fixes for bug 1023 2013-03-29 15:05:07 +01:00
Axel Uhl b7182c4553 fixed the first of the leaderboard tests broken by the change in behavior requested by the fix for bug 1023 2013-03-29 14:56:41 +01:00
Simon Pamies bd3999943b Adding information about java.io.tmpdir 2013-03-27 19:33:03 +01:00
Simon Pamies b28e8efe32 Somehow my eclipse did not save change related to socket count 2013-03-27 19:33:03 +01:00
Simon Pamies bc5ea8b84a Clarifying information provided by monitoring, also trying to get socket count right 2013-03-27 19:33:03 +01:00
fmittag 605332104e Merge remote-tracking branch 'origin/master' into leaderboard_json_doc_and_test 2013-03-22 17:24:38 +01:00
fmittag e3df4ea5dd added a json servlet for exporting wind data (similar to the showWind action from the AdminApp but with more parameters) 
some renaming of the json servlets
 2013-03-22 17:22:24 +01:00
Axel Uhl ab7fc9c519 Fixed by 1317: ConcurrentModificationException in SmartFutureCache unpropagating lock set 2013-03-22 15:09:42 +01:00
Axel Uhl 057a479586 Largely fixed bug 1023. While it doesn't show an additional column with the race count, LowPoint and HighPoint and all 
their derived schemes rank a competitor better if it has more races; all races of those schemes are now always
considered valid for total points score. The old implementation is still in AbstractScoringScheme in case someone
needs them in a specific scheme.
 2013-03-22 15:04:18 +01:00
Axel Uhl 98c14212aa Merge remote-tracking branch 'server/master' 2013-03-22 14:05:55 +01:00
Axel Uhl 0313c26a2d in LockUtil, consider the possibility that a thread from which locks are to be carried over doesn't hold any 2013-03-22 14:03:05 +01:00
fmittag 75751b7550 Merge remote-tracking branch 'origin/master' into leaderboard_json_doc_and_test 2013-03-22 11:51:13 +01:00
fmittag a36606226f added sample json files for all json exports 2013-03-22 11:47:02 +01:00
Axel Uhl eb7245c3dc added a comment on the minimum size of the thread pool for the SmartFutureCache 2013-03-22 11:26:52 +01:00
fmittag 8b8989dc3e fixed missing method renaming 2013-03-22 10:58:30 +01:00
fmittag e4a4219624 replaced EventDataJsonExportServlet accessible via /rc/events with more general EventsJsonExportServlet via /events 2013-03-22 10:54:21 +01:00
fmittag 4681e78b61 Merge remote-tracking branch 'origin/master' into leaderboard_json_doc_and_test 
Conflicts:
	java/com.sap.sailing.server.gateway/WEB-INF/web.xml
	java/com.sap.sailing.server/src/com/sap/sailing/server/impl/RacingEventServiceImpl.java
	java/com.sap.sailing.server/src/com/sap/sailing/server/operationaltransformation/AddDefaultRegatta.java
 2013-03-22 10:45:34 +01:00
Axel Uhl 3dfc8ca0e7 apply lock propagation also to write locks 2013-03-22 00:34:45 +01:00
Axel Uhl f9292e64c3 added lock propagation to LeaderboardDTOCache and SmartFutureCache 2013-03-22 00:28:23 +01:00
Axel Uhl 82b464eafe added comment to SmartFutureCache.triggerUpdate to explain the need for randomization in case of mass updates 2013-03-22 00:09:14 +01:00
Axel Uhl e73b325db7 Merge remote-tracking branch 'server/master' 2013-03-21 23:59:20 +01:00
Axel Uhl 91b684804b use a single ThreadPool for all SmartFutureCache instances, 3*availableProcessors in size; shuffle multi-cache-invalidations; 
when invalidating a SmartFutureCache for a number of keys, if this always happens in the same order, the first few keys
will be re-computed over and over again because cancelling them is not possible as they have already started, but a new
re-calculation needs to be triggered still. All later keys will be appended to the ordered queue but will keep getting
canceled because they haven't started their calculation yet. By shuffling the keys before triggering their invalidation,
all elements will fairly get re-calculated every now and then.
 2013-03-21 23:53:39 +01:00
Axel Uhl 9b3f696620 use a ConcurrentHashMap to avoid issues in LockUtil caused by concurrent access to the lockCounts map 2013-03-21 23:48:08 +01:00
Axel Uhl 40af74b11b trying to fix bug 1312 by allowing LeaderboardDTOCache to use stale results of CrossTrackErrorCache and ManeuverCache; 
The LeaderboardDTOCache is invalidated upon each GPS fix received; therefore, by and large, the LeaderboardDTOCache will
be invalidated each time the maneuver and cross track error caches will be invalidated. Only in very bizarre situations
such as the end of the race stopping all GPS fixes would the LeaderboardDTOCache keep some stale results for the very end
of the race. The likelihood seems sufficiently small to tolerate the risk.
 2013-03-21 22:12:00 +01:00
Axel Uhl 51ddeef2b8 working towards a solution against read-read locks when first reader waits for second reader (see bug 1313) 2013-03-21 21:22:01 +01:00
Armin Zamani d1f64c456d renamed "CourseData" to "CourseBase", "EventData" to "EventBase" and 
"SeriesData" to "SeriesBase", and their impl-classes accordingly,
renamed "CourseDataJsonSerializer" to "CourseBaseJsonSerializer"
 2013-03-21 18:30:30 +01:00
Armin Zamani c97edeb46b Renamed AddToRaceLogJsonExportServlet to 
AddEntryToRaceLogJsonPostServlet, renamed path /rc/leaderboards to
/rc/racegroups
 2013-03-21 18:28:13 +01:00
fmittag 887b2ec5dd added a json export servlet for sailing events 2013-03-21 18:05:48 +01:00
Axel Uhl feaea37a0c improved trace output for locking 2013-03-21 17:43:23 +01:00
fmittag 9e0b09fab7 added test cases for json export of a leaderboard with "Live" and "Final" result state 2013-03-21 14:37:58 +01:00
fmittag bd9ef9b963 Merge remote-tracking branch 'origin/master' 2013-03-20 19:16:06 +01:00
fmittag 9a3319b5f1 fixed a bug with the replication of default regattas (id replication was broken) 
added a test case for replicating default regattas
 2013-03-20 19:12:15 +01:00
fmittag d48872b809 used a different id for regatta creation due to bug 1311 2013-03-20 17:35:37 +01:00
Axel Uhl 7211f60040 fixed use of LowPointWinnerGetsZero scoring scheme 2013-03-20 16:40:54 +01:00
fmittag 8aa04590ff Merge remote-tracking branch 'origin/master' into leaderboard_json_doc_and_test 2013-03-20 15:33:54 +01:00
fmittag f49cccea25 added basic json export tests for leaderboardGroups and a leaderboard 2013-03-20 15:33:32 +01:00
fmittag 5f66e093ce Merge remote-tracking branch 'origin/master' 2013-03-20 14:16:49 +01:00
fmittag de37ac1a0f fixed NPE in boatclassimagedata when no compatible boat classes are available 2013-03-20 14:13:45 +01:00
fmittag 123e58bc3d initial commit for documentation of gateway web services (json exports) 2013-03-20 12:09:45 +01:00
fmittag 36c60fb4b4 first test case for json exports 2013-03-20 12:08:58 +01:00
fmittag 26da06454e cleanup of method parameter name during regatta creation (baseRegattaName instead of baseEventName) 2013-03-20 12:07:47 +01:00
Axel Uhl 8609ff84d1 fixed competitor ID problem for SwissTiming; pesistence layer assumed it could put a Serializable ID into a DBObject which is not the case; 
More generally, this raises the question where else we rely on Serializable going into MongoDB. Opened bug 1311
 2013-03-20 11:28:20 +01:00
Axel Uhl 861913a009 fixing issue with selection of XRR document if multiple matches with different time points exist 2013-03-20 10:44:12 +01:00
Axel Uhl 18d31011cd added NOAA importer launch configurations for declination import into resources/ 2013-03-20 00:33:50 +01:00
Axel Uhl 64fa5ab77e adjusted XRR import test verdicts after fixing the event name / boat class name mix-up 2013-03-19 22:42:27 +01:00
Axel Uhl 807ca61b7b handle non-existing results in score correction document 2013-03-19 21:41:43 +01:00
Axel Uhl 937d54b3ed added XRR import bundle to launch configs; fixed mix-up of boat class name and event name in XRR result import 2013-03-19 21:11:44 +01:00