Commit Graph
67 Commits
Author SHA1 Message Date
Axel Uhl f458607942 some comment changes 2015-07-08 00:05:36 +02:00
Axel Uhl a032146c63 use setCurrentlyFillingFromInitialLoadOrApplyingOperationReceivedFromMaster in finally clause and not only in exception handler 2015-07-07 11:16:08 +02:00
Axel Uhl f11fdfe75f added logging for operations replicated from replica to master 2015-07-07 11:15:45 +02:00
Axel Uhl 6fd6c5a762 solving bug 3015 according to comment #12; only the replicables requested by the system property will have initialLoad requested and will be waited for 2015-06-29 13:44:54 +02:00
Axel Uhl 6d2f2ae3d0 improved logging output for replicables start-up 2015-06-28 14:07:40 +02:00
Axel Uhl 22d8f8a65b make sure that the replicables list used for URL construction is the same as the one for de-serializing the replicables 2015-06-28 13:00:44 +02:00
Axel Uhl 03074f3f28 resume replicator to apply queued messages even if there has been an exception for one of the replicables 2015-06-28 12:14:17 +02:00
Axel Uhl e991a9b53b always use the same order of replicables in startToReplicateFrom for all aspects of iterating over them 2015-06-28 12:06:04 +02:00
Axel Uhl f170a0a13a working towards an improvement for initial load serialization consistency: flush ObjectOutputStream before opening the next on the same underlying OutputStream 2015-06-25 19:15:11 +02:00
Axel Uhl 75b04d8048 Merge branch 'master' into ubilabs--tracking_app--develop 2015-03-30 22:09:02 +02:00
Axel Uhl 9c5978ba4f notify waiters if ReplicationReceiver is stopped for any reason, setting stopped to true; example: consumer closed exception 2015-03-30 20:51:12 +02:00
Axel Uhl 824f975ba8 renamed isQueueEmpty to isQueueEmptyOrStopped; don't set thread context class loader in ReplicationReceiver: it's set in the replicable itself 2015-03-30 20:28:36 +02:00
Axel Uhl 9001649b99 fixed waiting on Replicator's queue to become empty; this avoids the AlreadyClosedExceptions, it seems, but the testWindSourcesToExcludeReplication test still fails locally 2015-03-30 11:50:48 +02:00
Axel Uhl b8ef26800f improved replication test logging; improved Javadoc for RaceTracker.stop(...) 2015-03-30 10:44:08 +02:00
Axel Uhl 21d5d33f78 fixing re-opened bug 2707, now also for replication-related de-serialization in RacingEventServiceImpl, using the same class loader 2015-03-29 18:00:07 +02:00
Fredrik Teschke 9c375a3ae4 Merge remote-tracking branch 'origin/master' into android_apps_no_rcapp_changes
Conflicts:
	.gitignore
	java/com.sap.sailing.server/SailingServer (No Proxy).launch
	mobile/com.sap.sailing.android.shared/.classpath
	mobile/com.sap.sailing.android.tracking.app/src/com/sap/sailing/android/tracking/app/ui/activities/LaunchActivity.java
2015-02-13 11:16:50 +01:00
Axel Uhl ba65988793 Merge remote-tracking branch 'server/master'
Conflicts:
	java/com.sap.sse.security.ui/Security UI sdm.launch
2015-02-06 05:58:55 +01:00
Axel Uhl f0ecc3f789 leaving a FIXME comment regarding bug 2600 2015-02-05 18:28:04 +01:00
Axel Uhl df332aa4ba use Lambda syntax for a simple Runnable 2015-02-05 16:00:23 +01:00
Fredrik Teschke 4508934766 remove com.sap.sse.{gateway, osgi} by moving the few classes into com.sap.sse 2015-02-05 09:28:38 +01:00
Axel Uhl 1b048d23e0 improved logging in case of exceptions during initial load 2015-02-05 03:34:11 +01:00
Axel Uhl 9bc84d1079 log and send error to client asking for initial load for a replicable the master doesn't know 2015-02-04 16:53:18 +01:00
Peter Siegmund c580c10f89 Merge branch 'bug2572' into android_apps 2015-01-14 17:02:18 +01:00
Axel Uhl 7f7eb10256 move ObjectInputStreamResolvingAgainstCache to com.sap.sse and let com.sap.sse depend on com.sap.sse.common which has the IsManaagedByCache interface 2015-01-14 13:54:45 +01:00
Axel Uhl 24bba04589 moved ObjectInputStreamResolvingAgainstCache to com.sap.sse.common so that com.sap.sailing.domain does not require a dependency on com.sap.sse.replication anymore 2015-01-14 10:59:05 +01:00
Axel Uhl 7181b2ecf9 fixing grave bug with replication: replicator was being resumed after receiving *first* initial load, but we need to wait until after the *last* initial load 2014-12-12 07:41:20 +01:00
Axel Uhl 64451c2702 fixing bug 2493; introduced a ThreadLocal telling whether replica is receiving initial load or running a master operation; not firing back to master if this flag is set 2014-12-11 09:57:18 +01:00
Axel Uhl 70647f1afc added better Javadoc for operation ID decoration during transitive replication 2014-12-01 09:59:20 +01:00
Axel Uhl 3e456170bf fixed issue described by comment #5 in bug 2408; the ID of operations sent from replica to master is stored in thread local and attached to operations sent out on behalf of this 2014-11-30 01:47:22 +01:00
Axel Uhl 8e7ab908e7 replicate operations initiated on replica to master and ignore them when they come back to that replica; see bug 2408 2014-11-29 23:50:06 +01:00
Axel Uhl 22d9ecdd3b working on bug 2408, adding an implementation that has the replica side of serializing an operation to the master through an HTTP request 2014-11-28 18:12:09 +01:00
Axel Uhl 3b499d8492 working towards a solution for bug 2480 by adding OperationWithResult.isRequiresExplicitTransitiveReplication and respecting it in replicate(...) 2014-11-28 16:39:08 +01:00
Axel Uhl 8ebab4ca50 basic authentication and session replication is in place; no longer use EhCache as it seems very tricky to replicate; still issues with details of session authentication replication 2014-11-24 17:56:43 +01:00
Axel Uhl 051f867a29 moved setting of context class loader for initial load deserialization into abstract implementation 2014-11-24 14:30:37 +01:00
Axel Uhl b36b2e146f added missing pom.xml stuff for replication.testsupport 2014-11-21 12:28:03 +01:00
Axel Uhl 801946302f a first compileable version where SecurityService is a replicable 2014-11-20 21:20:08 +01:00
Axel Uhl 355239c8b1 first steps towards making SecurityService a Replicable; introduced operation base interface; register SecurityService as a Replicable; introduced ObjectInputStream / Cache 2014-11-20 17:39:58 +01:00
Axel Uhl 5439757b13 fixed replication operation queueing in Replicator; adjusted PrematureOperationReceiptTest to fail with unfixed queueing 2014-11-20 11:25:06 +01:00
Axel Uhl 6956567574 refactored AbstractRaceChangeListener so that subclasses can provide a defaultAction() implementation; in preparation of solving bug 2473. 2014-11-20 09:58:11 +01:00
Axel Uhl ffea760069 avoid premature replication listener registration when no replica is currently registered 2014-11-17 14:58:17 +01:00
Axel Uhl 28e572a94c fixed synchronization with waiting for Replicable instances 2014-11-17 00:12:36 +01:00
Axel Uhl 9df49648f5 addressing bug2409 comment #9 by changing semantics of -Dreplicate.on.start from boolean to list of fully-qualified Replicable class names 2014-11-16 23:21:34 +01:00
Axel Uhl 6bd64b7e7d use a (currently fixed) list of Replicable IDs to wait for before auto-launching replication 2014-11-16 23:04:35 +01:00
Axel Uhl 417dbd0015 slight improvements for replication master descriptor toString(); handle missing replicables a bit better 2014-11-16 19:05:45 +01:00
Axel Uhl 15ff17798a set thread context class loader for master data import to ensure ObjectInputStream can resolve all classes as needed 2014-11-14 15:48:37 +01:00
Axel Uhl f1e23da6b1 consolidate imports on Replicable, avoiding warnings 2014-11-14 13:06:54 +01:00
Axel Uhl 59803c533c use Replicable.writeOperation in ReplicationServiceImpl 2014-11-14 13:06:08 +01:00
Axel Uhl 8b195d5fed factoring StreamingOutput classes in MasterDataResource 2014-11-14 13:03:51 +01:00
Axel Uhl 29db1e0858 fixed initial load serialization for replication tests 2014-11-14 00:21:16 +01:00
Axel Uhl a808d78d14 all compile errors removed in Eclipse workspace for bug2409; now it's time for testing... 2014-11-13 23:51:45 +01:00