Patrik Nordwall
79d8ec87fc
Delay should not pull when buffer is full with Backpressure strategy, #21334
...
Additionally
* nano time conversion and calculation fix
* refactoring of Delay.onPush
2016-09-13 16:10:49 +02:00
Patrik Nordwall
86c15c04e3
Merge pull request #21453 from johanandren/wip-20724-increase-LeveldbAtLeastOnceDeliverySpec-johanandren
...
Verbose logging in LeveldbAtLeastOnceDeliverySpec
2016-09-13 15:52:21 +02:00
gosubpl
9b73fefdce
=htc don't encode known to be empty entities ( #21393 ) ( #21396 )
...
* =htc don't encode known to be empty entities (#21393 )
* =htc added missing headers in the improved marshaller (#21393 )
2016-09-13 14:32:02 +02:00
drewhk
ec50bd1441
Merge pull request #21447 from drewhk/wip-21446-harden-event-chasing-drewhk
...
#21446 : Completion events must not be swallowed if chasing
2016-09-13 12:17:02 +02:00
Johan Andrén
74162ddc9f
Verbose logging in LeveldbAtLeastOnceDeliverySpec to figure out #20724
2016-09-13 11:46:18 +02:00
Johan Andrén
e65e63e8c1
=per Higher recovery timeout in PersistentActorRecoveryTimeoutSpec to not fail on slow/gc #20728 ( #21445 )
2016-09-13 10:51:40 +02:00
Konrad Malawski
3b57947b1f
=art remove addressed todo in compressions
2016-09-13 10:49:19 +02:00
Łukasz Dubiel
211537f485
=art Count min sketch benchmark and performance improvments. ( #21037 )
2016-09-13 10:48:37 +02:00
Endre Sándor Varga
b2f0ca6750
#21446 : Completion events must not be swallowed if chasing
2016-09-13 10:47:08 +02:00
Matthias Lüneberg
a88c8a0bbe
Add javadsl FormData.create(Iterable) overload #21303 ( #21344 )
2016-09-13 10:19:37 +02:00
drewhk
47433bc65c
Fix initialization order in BroadcastHub caused by an otherwise innocent race #21362
2016-09-13 09:02:53 +02:00
Patrik Nordwall
bf151e9793
don't quarantine back, #21450
...
* Don't quarantine the other system when receiving the Quarantined message,
since that will result cluster member removal and can result in
forming two separate clusters (cluster split).
* Instead, the downing strategy should act on ThisActorSystemQuarantinedEvent, e.g.
use it as a STONITH signal.
2016-09-13 08:01:58 +02:00
Patrik Nordwall
acafe80cf1
rewrite TestStage to use thread-safe shared state, #21431
...
* The previous approach was based on sending the
test commands to the active stages themselves and let
them keep track of the state.
* The problem with that is that Association/OutboundTestStage
that is created afterwards will not have the right state.
Similar problems can occur for restarts.
* Instead using thread-safe mutable state that is
updated directly and used by all test stages.
2016-09-12 19:51:05 +02:00
Richard Imaoka
98bfbbb4b3
Add documentation for ReplayFilter #20301
2016-09-12 16:18:06 +02:00
Johannes Rudolph
a6ad4659d6
Merge pull request #21444 from jrudolph/cleanup-header-builder
2016-09-12 15:42:59 +02:00
Johannes Rudolph
0e253b4208
=rem,act move direct buffer cleaning util method to akka.io
2016-09-12 14:25:30 +02:00
Johannes Rudolph
bffbef0caa
=rem remove test-only HeaderBuilder.bothWays
2016-09-12 14:20:37 +02:00
Johan Andrén
75bbd7d00b
-doc Remove stale other language bindings doc page #21426 ( #21432 )
2016-09-12 11:59:38 +02:00
Richard S. Imaoka
ae084083f6
More descriptive errors from ReplayFilter ( Fixes #20394 )
2016-09-10 02:05:36 +09:00
Johan Andrén
3502f0d72f
One more missed canonical.port in cluster tests ( #21428 )
2016-09-09 18:12:35 +02:00
Johan Andrén
b0e03058b9
Port and hostname config path was changed, cluster tests didn't get the change ( #21427 )
2016-09-09 17:55:02 +02:00
Johan Andrén
fa1d6d6f19
Disable ClusterDeathWatchSpec for now ( #21421 )
2016-09-09 17:54:13 +02:00
Patrik Nordwall
e2ea9fc46a
Merge pull request #21414 from akka/wip-21403-banned-resolveActorRef-patriknw
...
try delayed retry resolve of remote deployed refs only once, #21403
2016-09-09 15:42:53 +02:00
Patrik Nordwall
d8bb0ef476
Merge pull request #21406 from akka/wip-21371-prio-patriknw
...
No ack delivery for prio messages, #21371
2016-09-09 15:41:54 +02:00
Patrik Nordwall
7513617070
Merge pull request #21417 from drewhk/wip-20623-cleanup-aeron-files-drewhk
...
#20623 Make sure external (mapped) resources are properly cleaned on shutdown
2016-09-09 15:23:13 +02:00
Patrik Nordwall
4e34e4373c
Merge pull request #21418 from akka/wip-sync-2.4.10-patriknw
...
Sync Artery with 2.4.10
2016-09-09 15:11:12 +02:00
Patrik Nordwall
1584c52190
handle longer network partitions, #21399
...
* system messages in flight should not trigger premature quarantine
in case of longer network partitions, therefore we keep the control
stream alive
* add give-up-system-message-after property that is used by both
SystemMessageDelivery and AeronSink in the control stream
* also unwrap SystemMessageEnvelope in RemoteDeadLetterActorRef
* skip sending control messages after shutdown, can be triggered
by scheduled compression advertisment
2016-09-09 14:35:50 +02:00
Patrik Nordwall
cd4a31e74d
No ack delivery for prio messages, #21371
...
* and send prio messages enclosed in actor selection
over the control stream
2016-09-09 14:35:50 +02:00
Endre Sándor Varga
0d77034adc
20623 Make sure external (mapped) resources are properly cleaned on shutdown
2016-09-09 14:29:04 +02:00
Patrik Nordwall
ae11fb3b45
Merge pull request #21413 from akka/wip-21339-enable-misc-serial-patriknw
...
enable misc serializers by default for Artery, #21339
2016-09-09 14:29:02 +02:00
Patrik Nordwall
a5f693c2bc
Merge pull request #21412 from akka/wip-21397-MergeHub-log-patriknw
...
add recover in front of MergeHub, to avoid logging, #21397
2016-09-09 14:27:53 +02:00
Patrik Nordwall
e8ce261faf
Merge branch 'master' into wip-sync-2.4.10-patriknw
2016-09-09 14:12:16 +02:00
Richard Imaoka
0f2da7b26b
Add instruction for akka-samples ( Fixes #21415 ) ( #21416 )
2016-09-09 13:15:33 +02:00
Martynas Mickevičius
1ce7d7d7e9
#20946 Add bind address ( #21404 )
2016-09-09 12:46:50 +02:00
drewhk
e49b11607a
Merge pull request #21353 from drewhk/wip-21206-safer-afr-close-drewhk
...
#21206 : Improve AFR resilience to write-after-close
2016-09-09 10:20:16 +02:00
Patrik Nordwall
7b665f705e
try delayed retry resolve of remote deployed refs only once, #21403
...
* If the retried resolve isn't successful the ref is banned and
we will not do the delayed retry resolve again. The reason for that is
if many messages are sent to such dead refs the resolve process will slow
down other messages.
2016-09-09 10:15:12 +02:00
Patrik Nordwall
97e0628173
enable misc serializers by default for Artery, #21339
...
* placed them in a new section additional-serialization-bindings,
which is included by default when Artery is enabled
* can also be enabled with enable-additional-serialization-bindings
flag to simplify usage with old remoting
* added a JavaSerializable marker trait that is bound to JavaSerializer
in testkit, this can be used in tests so that we eventually can run
tests without the java.io.Serializable binding
2016-09-09 09:01:15 +02:00
Patrik Nordwall
494ccc00dc
add recover in front of MergeHub, to avoid logging, #21397
2016-09-08 19:34:18 +02:00
Patrik Nordwall
02de58392a
Merge pull request #21398 from akka/wip-fixme-patriknw
...
fix a bunch of FIXMEs in Artery
2016-09-08 19:15:43 +02:00
Patrik Nordwall
fea18a9a1b
mute expected exception in RemoteDeathWatchSpec
2016-09-08 18:00:37 +02:00
Patrik Nordwall
c74ece9573
require power of two
2016-09-08 18:00:37 +02:00
Patrik Nordwall
74a8bb3a00
flight recorder event for send queue overflow
2016-09-08 18:00:37 +02:00
Patrik Nordwall
3b7a7dfa59
add reason param to quarantine method
2016-09-08 18:00:37 +02:00
Patrik Nordwall
9a7d79c882
size of outboundEnvelopePool
2016-09-08 18:00:37 +02:00
Patrik Nordwall
8756ffd75c
handle Aeron Publication.CLOSED
2016-09-08 18:00:37 +02:00
Patrik Nordwall
3c779cebd4
config of send queues
2016-09-08 18:00:37 +02:00
Patrik Nordwall
ebd1883df5
remove or reword obsolete fixme
2016-09-08 18:00:37 +02:00
Johannes Rudolph
529b10a9f8
=rem #21202 a first set of protobuf serializations for artery messages ( #21380 )
2016-09-08 17:58:25 +02:00
Johan Andrén
b127f1fdcf
Rename protocol from "artery" to "akka" ( #21405 )
2016-09-08 16:12:29 +02:00
Endre Sándor Varga
8513ac9752
#21206 : Improve AFR resilience to write-after-close
2016-09-08 14:08:18 +02:00