* Started Implementing the ServerBasedlocking. There now is a container that can contain multiple query snippets. It now has to setup the necessary calls to the Servers
* Added backwards linking of QueryEngines, sth. DBServers can contact their Coordinators.
* Added LogTopic AQL
* Made AccessMode::Type Hashable
* Created a Mapping Server => LockLevel => Shard and createad a JSON object containing the Lock information for a complete AQL query per server
* Added code to build coordinator engines
* Finished with first draft of Coordinator-side of new DBServer based locking.
* Added a _api/aql/setup route that creates and locks all snippets/collections for one DBServer in a single go
* Fixed some Coordinator parts
* Index node now gracefully reports if it could not find it's collection when created from vpack. Otherwise it just hardly crashed...
* Modified the Coordinator Snippet collector to be able to handle subqueries properly.
* Started adding GraphNode handling. WIP. Need to deploy engines properly. Coordinator crashes on Graph tests
* Fixed compiler errors
* WIP: EngineInfoContainer
* Separated the EngineInfoContainers for Coordinator and DBServer into different files. They diverged more than anticipated
* Added forgotten files. THe DBServer container now creates the TraverserEngine Mapping and moves it into the Infos. They are not keeping it yet and need to add it to the message as well.
* The DBServer engine infos now persist the TraverserEngine infos. Need to add them to messages though.
* The new aql exec-engine now sends out traverserEngines as well
* Formatting and adding DEBUG level output
* Made the RestAQLHandler aware of the TraverserEngineRegistry. Also created the engines now. Return format changed server-side coordinator side needs fix.
* Adapted the Coordinator side for the DBServer based Shard Locking
* The DBServer based Locking now honors restrictions to certain shards
* Fixed a strange double lock bug in the new AQL Server based locking technique. Add some DEBUG output
* Fixed usage of MAINTAINERMODE macro. The assertion was never active
* Added TestCase for ContainerCoordinatorTest to cmake
* Added -DTEST_VIRTUAL to CMAKE. This is used to define virtual functions for mocking ONLY on test-builds.
* Fixed usage of ENABLE_MAINTAINER_MODE ifdef. CLANG format
* On non-enterprise builds ENTERPRISE_VERT defaults to TEST_VIRTUAL => virtual in test else non-virtual
* Added TEST_VIRTUAL to ExecutionEngine, Query and QueryRegistry
* Added first testcase for EngineInfoContainerCoordinator not yet ready.
* Mode CreateBlock a member function of engine, we have the engine in our hands anyways no need to make it static. Included some more TEST_VIRTUAL functions.
* Fixed clang/MacOs compile error. Added some more TEST_VIRTUAL declarations
* Finally fixed the first buildEngines UnitTest \o/
* Added a unit-test for backward linking of dependencies in CoordinatorPlanner
* Added multi-snippet test for EngineInfoContainerCoordinator
* Removed QueryRegistry.h from central header files and replaced by a forward declartion.
* Added a createBlocks method on the ExecutionEngine. It should be responsible to create all those blocks at once. Adapted the UnitTests as well. Not included Tests for the new createBlocks functionality. Need to mock the options feature first
* Added another test that Coordinator Snippets of queries can be created correctly
* Fixed Coordinator-site cleanup of QueryRegistry, if any of the query creations fails with error, incl UnitTest.
* Added first test for RestAqlHandler::setup. It does only test the setup and gives prepartion for real testing.
* Added a assertion of http return code. Still no creation of queries is tested. Requires a huge amount of mocking.
* fix some deadlocks found by evil lock manager (tm)
* fix duplicate lock
* fix indentation
* ensure proper lock dependencies
* fix lock acquisition
* removed useless comment
* do not lock twice
* create either a V8 transaction context or a standalone transaction context, depending on if we are called from within V8 or not
* AQL micro optimizations
* use explicit constructor
* only use V8DealerFeature's ConditionLocker for acquiring a free V8 context
entering and exiting the selected context is then done later on without having to hold the ConditionLocker
* remove some recursive locks
* Disable custom deadlock detection when Thread Sanitizer is enabled
* Changing ifdef's
* grr
* broke gcc
* Using atomic for ApplicationServer::_server
* fix premature unlock
* add some asserts
* honor collection locking in cluster
* yet one more lock fix
* removed assertion
* Allow the clustercomm to send nolock headers on count. This is used form within AQL
* IsLocked on transactions will now always yield true IF LOCK_NEVER is set. We simply assume someone else holds the lock for us. Also LOCK_NEVER is now set on collection/count if noLock header is send.
* Moved the flag if collections need to be locked into the TraverserEngines.
* Added enterprise-satellite hooks in EngineInfoContainerDBServer
* Removed now obsolete code
* Replaced throwing of Exception by an ResultObject
* Added some more tests and moved adding snippet to query engine more to the outside.
* Added the AQL result type
* Make the branch compile again
* Register WITH collections for Graphs in the new Collector.
* Fixed test code for failing query clone. Idea was to once clone successfully and second time to fail, we verify that first clone is cleaned up properly. However test failed on first clone...
* Removed a double builderClose
* Added Changelog entry
* Removed empty if
* Removed obsolete todo
* Properly initialize the AqlResult with nullptr on error case
* Updated comment
* Simplified Assertion
* Removed debug output object...
* Added additional catch case for std::exception to get some more error info
* Clarified evaluation order for move case
* Added Explicit
* Fixed cleanup of Coordinator if Registry fails to insert query.
* Allow to use other locks than Read/Write for AQL collections. Not yet in API.
* Updated Comments for other Locks on DBSide. Adapted Destruction CatchTests
* Fixed double builderClose and removed unnecessary double commits
* Added a comment to clarify the state
* Moved error output to trace. Leftover from debugging
* Added some tests for complex subquery patterns
* Added a 'fireAndForgetRequests' methods to cluster comm which allows to send out a bunch of messages but does not wait for their results
* Properly cleanup leftovers of queries if the instantiation step already failed
* Added code comment for fireAndForget
* Added indexes to subquery test to make the plan a bit easier
* The cleanup on DBServerEngines in error case now also cleans up traverser engines.
* Removed unnecessary includes
* Removed debug logging
* Fixed hidden merge conflict
For example, the following log level now does not cause a warning anymore
v8 = trace
previously the log levels had to be specified without any whitespace around
* remove some non-unused V8 persistents
* do not throw that many bogus assertions
* do not rely on server role being defined
* slightly better debug output for V8 context debugging
* fix collection ids in inventory response
* simplify bootstrap a bit
* slightly better error handling
* make elapsed time a queryable value
* use less memory for stub collections
* added assertions that will always make sense
* added assertions
* do not garbage-collect while waiting
* less copying of parameters
* do not show "load indexes into memory" buttons for mmfiles engine
as all indexes are in memory anyway
* when a collection is truncated via the web interface, flush the WAL and rotate all active journals
this will make close all open journals on leader and followers and make them subject to compaction opportunities
* fix invalid server id values being passed from web interface to backend
* introduce afterTruncate method for indexes
* added test case for issue #3447
* updated CHANGELOG
* don't warn about replicationFactor for system collections
* check that the queries actually use the geo index and not some other index
* properly report error in web interface
* fix some internals checks that made truncate fail for bigger collections in maintainer mode
* also run a compact() operation after a serious truncate
in order to make iteration over the truncated range much faster
when the collection is next accessed
* increase default maximum number of V8 contexts to at least 16
* improve error reporting in RestAqlHandler
* added logging about indexes
* added some safety checks to the logger
* slighty better error messages
* fix location header for SSL
* fix error message
* try to make tests more deterministic
* change error code from TRI_ERROR_INTERNAL (which we want to avoid) to TRI_ERROR_FAILED
* do not use V8 variant of AQL functions in early optimization stage when a C++ variant is available
* additionally, simplify AQL function definitions and aliases
* warn when more than 90% of max mappings are in use
* added C++ variant of replication catchup
* added `--log.role` option
* updated CHANGELOG
* removed non-existing scheduler.threads option from config
* removed useless __FILE__, __LINE__ invocations
* updated CHANGELOG
* allow a priority V8 context
* remove TRI_CORE_MEM_ZONE
* try to fix Windows errors & warnings
* cleanup
* removed memory zones altogether
* exclude system collections from collection tests
* removed delete call
* cleanup
* lower cpu activity of log thread too
* fix log messages
* do not enter threads into unordered_set, as it is unneeded
* do not compile in calls to disabled plan cache
* moved AQL regex cache from thread local variables to a class of its own
* more sensible thread creation and destruction
* remove API docs for endpoints that were deleted in 3.0
* remove unused variable
* set all scalar values in the ctor, so check tools do not complain
* move Result objects
* prevent potential nullptr access
* remove ambiguous boolean parameter with no clear meaning
* make code check tools happy
* don't allow copying cache Findings, allow moving Results
* the debug functionality for counting RocksDB key ranges was broken
this produced sporadic test failures
* remove superfluous iostream include
* fix range checks
* don't leak if an error happens during database removal
* try to simplify Result a bit
* try to deletion checks a bit more robust
* fix buffer overruns in linenoise for long input lines
* don't make historian repeatedly print the same error messages that nothing can be done about
* make the implementations of the logging operator<<s not throw exceptions, so that logging does throw exceptions as an unintended side effect
* update CHANGELOG
* improve error message
* don't copy strings, but pass them by const reference
Squashed commit of the following:
commit bc0472d212b2efef0d7b1b8a36f2b43e8432ba86
Merge: eb47631 5f87674
Author: Michael Hackstein <michael@arangodb.com>
Date: Mon May 29 09:39:31 2017 +0200
Merge branch 'devel' of github.com:arangodb/arangodb into feature/rocks-index-warmup
commit eb476310b0ca6165be10b37b960933886c2757f5
Merge: d725f21 32149d2
Author: Michael Hackstein <michael@arangodb.com>
Date: Mon May 29 09:00:41 2017 +0200
Merge branch 'devel' of github.com:arangodb/arangodb into feature/rocks-index-warmup
commit d725f21f7d61f7c79ba31fc0338881e35f4d8f48
Author: Michael Hackstein <michael@arangodb.com>
Date: Fri May 26 15:53:23 2017 +0200
Activated collection warmup in Cluster Mode. Added a test for collection warmup (SingleServerOnly)
commit 132bf4f9565b0dcf4ec9f84e93897b482a67ce7f
Author: Michael Hackstein <michael@arangodb.com>
Date: Fri May 26 12:46:20 2017 +0200
Implemented client-side warmup. It can now be triggered via http and arangosh
commit 78ea449dff86118814a2f87bdb59dc16544d92b6
Author: Michael Hackstein <michael@arangodb.com>
Date: Fri May 26 11:25:40 2017 +0200
Fixed assertion.
commit dae80f6277dde1a52eadda506858cc36e235bd55
Author: Michael Hackstein <michael@arangodb.com>
Date: Fri May 26 11:08:46 2017 +0200
Improved the CuckooIndexEstimator computeEstimate function. It is now much better on collections with many different values
commit 7abf57876511ba369d7a577e1995d4575e98c7c8
Author: Michael Hackstein <michael@arangodb.com>
Date: Fri May 26 11:06:43 2017 +0200
Edge index warmup will now first check if a document is stored and if not will insert it. Furthermore it resizes the cache with an estimate of how many documents will be most likely inserted
commit 890d8ad4cdfd155249f060fedd5c798b9531d556
Author: Michael Hackstein <michael@arangodb.com>
Date: Fri May 26 11:04:47 2017 +0200
Adjusted thresholds in transactional cache. Wastly increased the amount of tries to get the lock. Lowered the fill-grade boundaries
commit 60972ab7151a3acb78e1aae4149de11d0da7aceb
Author: Michael Hackstein <michael@arangodb.com>
Date: Fri May 26 10:45:38 2017 +0200
Added new Logtpopic CACHE that should be used to log important information regarding caches, like debug ouput on evictions / resizing, or memory-pressure resizing going on
commit 2dfacb9aef6e3fde169032514baca386786d059c
Author: Michael Hackstein <michael@arangodb.com>
Date: Fri May 26 10:44:21 2017 +0200
Fixed an assertion in index iterator. It expected _from/_to to be cached but we modified to cache entire document
commit f05beccef65853c30eda1859a39c28ea2295bc71
Merge: 13c6abf 5c46430
Author: Michael Hackstein <michael@arangodb.com>
Date: Fri May 26 07:56:58 2017 +0200
Merge branch 'devel' of github.com:arangodb/arangodb into feature/rocks-index-warmup
commit 13c6abfbc2ddb451a689011110411df68247435e
Author: Michael Hackstein <michael@arangodb.com>
Date: Wed May 24 09:52:01 2017 +0200
Added a getExtra feature for the edgeIndex. It now returns the complete edge as extra (temporary). Modified the EdgeCache to cache token and edge document (temporary). Added a warmup Function to collections that is used to warmup the index caches on demand.
Jan
Jan Christoph Uhde
Simon
jsteemann
Dan Larkin-York
Michael Hackstein
Andrey Abramov
Max Neunhöffer
Manuel B
Frank Celler
Dan Larkin
hkernbach
Jan Christoph Uhde