raboof/pekko
1
0
Fork 0
Code Issues Pull requests Projects Releases 6 Packages Wiki Activity Actions 11
25018 commits 1 branch 31 tags 268 MiB
Commit graph

25018 commits

This branch
This branch
All branches
Author SHA1 Message Date
Johan Andrén
aa874d610f Require set of remembered entities to start to be nonEmpty 2020-06-11 16:58:41 +02:00
Johan Andrén
5e222214b6 Debug log for stashing while awaiting lease 2020-06-11 16:57:32 +02:00
Johan Andrén
c5aa918732 Language fix and don't showcase leveldb 2020-06-11 16:07:50 +02:00
Johan Andrén
25a1daa0d2 Majority plus in docs 2020-06-11 16:05:36 +02:00
Johan Andrén
a0fa284398 Deprecation version updated 2020-06-11 16:04:23 +02:00
Johan Andrén
c9e153a47f Tune test logs back to info 2020-06-11 16:03:22 +02:00
Johan Andrén
4a4a6065c1 Move MiMa excludes to correct version 2020-06-11 16:02:07 +02:00
Johan Andrén
a31260e8fc More private and @InternalApi sprinkling on top 2020-06-11 16:00:43 +02:00
Johan Andrén
1e811eebbc Avoid allocation on entity lookup 2020-06-11 15:58:33 +02:00
Johan Andrén
55311cf914 RememberingStartNoAck -> RememberingStart.empty 2020-06-11 15:56:56 +02:00
Johan Andrén
33b34da36c Union and diff for sets are more performant 2020-06-11 15:52:46 +02:00
Johan Andrén
9c7f16a4db Re-use logic for allocating shard home for remembered entities 2020-06-11 15:46:03 +02:00
Johan Andrén
d2dd80fe17 Some clarifying comments 2020-06-11 15:45:00 +02:00
Johan Andrén
078d7bd2fb
Handle some corner case states when receiving StartEntity (#29176) 2020-06-04 10:52:53 +02:00
Johan Andrén
1254595c7d
Remember entities starting separation of concerns (#29174) 
Extract starting strategy out of shard into RememberEntitiesStarter and do it through messages instead of futures.
 2020-06-04 10:43:08 +02:00
Johan Andrén
45d9d89fbe
Restore termination behavior for non-remember entities (#29175) 
... and add testcoverage
 2020-06-04 10:27:20 +02:00
Johan Andrén
fb39ac1a9c
Tests for EventSourced remember entities shard store (#29177) 
* Tests for EventSourced remember entities shard store
* Move the migration event adapter to a better place and mention in docs.
 2020-06-04 09:54:28 +02:00
Christopher Batey
eb923bbbe1
Move coordinator state and remembered entities next to each other (#29173) 
* Fix docs on how to passivate
* Remove 10000 limitation of remembered entities
 2020-06-03 11:55:51 +01:00
Johan Andrén
85e11c8941
Start entity ack not needed in shard (#29164) 
* The only sender of StartEntityAck to shard is RememberEntitiesStarter

That is not really needed, because the shard already knows if it started an entity, so we can drop that logic

* Separate command for the case where ids has been relocated by a new shard id extractor algo

* Store that entities were moved on shard id extractor change and cover with test

* Do'h. Can't mark entity as being remembered before checking if there are any entities being remembered

* Comment explaining test
 2020-06-03 09:40:33 +01:00
Johan Andrén
224fb1592d
Turn verbose logging into a config flag (#29168) 2020-06-03 08:42:17 +01:00
Johan Andrén
5a5468dd4a
DData shard store improvements (#29166) 
* Handle timeouts better wrt the timeout used in shard
* load all entities up front to potentially speed up GetRemembered response
 2020-06-03 09:02:32 +02:00
Johan Andrén
aa50c63965
Accumulate and ack all requests to start entity (#29165) 2020-06-02 16:30:16 +02:00
Johan Andrén
5643f7e194
Problems with StartEntityAck, StartEntity and the persistent test running with ddata remember entities (#29161) 2020-06-02 14:59:38 +02:00
Johan Andrén
13aed055fd
Publish Dropped message when sharding drops buffered messages (#29163) 2020-06-02 14:58:04 +02:00
Johan Andrén
f70d33145c
Merge pull request #29158 from johanandren/merge-master-2 
Merge Master into Remember Entities feature branch a second time
 2020-06-02 08:45:34 +02:00
Johan Andrén
5d42c929ef I gotta remove these empty files 2020-06-01 17:36:49 +02:00
Johan Andrén
e89b108765 Scalafmt wants an empty line after header 2020-06-01 15:51:32 +02:00
Johan Andrén
012140e814 Missing headers 2020-06-01 15:06:19 +02:00
Johan Andrén
5b5e3577e2 Some changes on top of changes from master 2020-06-01 14:30:28 +02:00
Johan Andrén
5bbf5a5a21 Merge branch 'master' again to get the latest sharding fixes 2020-06-01 12:57:51 +02:00
Johan Andrén
b9667fb6b3
Batch remember entity stops (#29149) 2020-06-01 11:03:03 +02:00
Patrik Nordwall
4425005435
Merge pull request #29144 from akka/wip-29085-spi-patriknw 
Telemetry SPI hooks for SBR decision, #29085
 2020-06-01 07:53:13 +02:00
Renato Cavalcanti
0969e9d5a3
Merge pull request #29028 from akka/actor-ref-provider-logging 
LocalActorRefProvider: added dedicated logger for deserialization failures
 2020-05-29 10:26:58 +02:00
Renato Cavalcanti
7770ea7ec9
Merge branch 'master' into actor-ref-provider-logging 2020-05-29 08:57:39 +02:00
Johannes Rudolph
24f2b2e617
stream: cosmetic improvements to FutureFlow (#29068) 2020-05-28 15:42:41 +02:00
Johannes Rudolph
842052d633
Merge pull request #29091 from jrudolph/28993-fix-keepAlive 
stream: make keepAlive/IdleInject more reliable against flukes
 2020-05-28 14:45:07 +02:00
Johannes Rudolph
8e349b4180
Merge pull request #29121 from jrudolph/29111-fix-tls-actor-hangs 2020-05-28 14:40:26 +02:00
Nicolas Vollmar
cfe4443d25
Extend backoff supervision with custom message handling #29082 (#29083) 2020-05-28 13:15:46 +02:00
Christopher Batey
0d1237fd44
Latency benchmarks for rememered entities + JFR events (#29103) 2020-05-28 09:35:27 +02:00
Patrik Nordwall
2a536d7065
Merge pull request #29140 from akka/wip-29085-log-patriknw 
Log markers for SBR, #29085
 2020-05-28 08:59:21 +02:00
Patrik Nordwall
cd9e9e960a Telemetry SPI hooks for SBR decision, #29085 2020-05-28 08:52:22 +02:00
Patrik Nordwall
95dd4b2021
handle thrown exceptions in Retry pattern in same way as failed Future (#29119) 2020-05-28 07:22:22 +01:00
Christopher Batey
870eef540a
Improve remember entities docs (#29136) 2020-05-27 15:50:55 +01:00
Johannes Rudolph
87c5844451
stream: avoid getting stuck when doUnwrap returns NEED_WRAP with outstanding data 
Refs #29111

This seems only to happen with TLS 1.3. In that case, remaining data in
`transportInBuffer` was left there instead of putting it back onto the
chopping block.

Then `doWrap` was run but `doUnwrap` was never called again because only
the chopping block was checked for outstanding data but not the buffer.
 2020-05-27 15:03:07 +02:00
Johannes Rudolph
9beb88fa44
stream: fix half-closed TLS connection spin-loop 
Refs #29110

TLSActor could get caught in a spin-loop on connection termination
because there was an implicit assumption that when inbound is closed
(peer has sent `close_notify`), this SSLEngine would also automatically
send a `close_notify` and close the connection.

Therefore, it would stay in `flushOutbound` pumping in a loop.

This is not true anymore with TLS 1.3, more accurately it can be
configured using `-Djdk.tls.acknowledgeCloseNotify` which is `false` by
default leading to half-open connections.

The solution is to not support half-open TLS connections for now and
consider a connection closed as soon as `isInboundClosed` and there's no
outstanding data.

(To support half-open connections, this fix would have to be reverted
and `flushOutbound` fixed accordingly.)
 2020-05-27 15:03:07 +02:00
Johannes Rudolph
e524a0d445
stream: run TlsSpec for TLSv1.3 on jdk >= 11 
Session renegotiation tests are disabled because TLS 1.3 does not
support renegotiation any more.
 2020-05-27 15:03:06 +02:00
Patrik Nordwall
b933fd0b16
Merge pull request #29137 from akka/wip-29131-region-termination-patriknw 
Don't allocate to terminated regions, #29131
 2020-05-27 14:30:47 +02:00
Patrik Nordwall
1472bd9e8c Log markers for SBR, #29085 2020-05-27 14:26:55 +02:00
Christopher Batey
b463a1adbd
Merge master into re feature branch (#29135) 
* Merge master into feature branch

* Formatting

* Remove redundant fixme

* Remove files that snuck in

* Fix backoff supervisor test
 2020-05-27 12:50:53 +01:00
Patrik Nordwall
95111955c4 Don't allocate to terminated regions, #29131 
* when using down-removal-margin it could allocate to an
  already terminated region
* the watch fix in PR #29092 solves this
* this is an "optimization" to avoid the regions that
  have been terminated
 2020-05-27 11:09:59 +02:00