aboutsummaryrefslogtreecommitdiffstats
path: root/persistence
Commit message (Collapse)AuthorAgeFilesLines
* Remove notion of node-specific reliability from C++ distribution codeTor Brede Vekterli2021-03-021-2/+1
| | | | | I have never seen this in use anywhere, and can find no code that ever sets it. Bonus is that node candidate trimming can be vastly simplified.
* - Change error handling so that both synchonous and asynchronous errors can ↵Henning Baldersheim2021-02-234-16/+23
| | | | | | | | be reported back from bucket executor. - Treat remapping as an error. - For lidspace compaction job iterator is reset and will be recreated on next invocation. - For bucketmove th ebucket is rechecked and either discarded or restarted.
* string.h does not need alloc.hHenning Baldersheim2021-02-212-0/+2
|
* Use a helper for bool -> vespalib::TrinaryHenning Baldersheim2021-02-191-7/+11
|
* Use a trinary to also handle temporary mismatching bucket used bits.Henning Baldersheim2021-02-193-14/+23
|
* - Reduce visibility of ClusterState and Distribution.Henning Baldersheim2021-02-192-10/+1
|
* Merge pull request #16495 from vespa-engine/arnej/add-size-literalsArne H Juul2021-02-161-14/+15
|\ | | | | add literals for sizes in KB, MB, and GB
| * use size literals in persistenceArne Juul2021-02-151-14/+15
| |
* | An explicit copy assignment is needed when an explicit copy constructor isTor Egge2021-02-162-0/+14
| | | | | | | | declared, to avoid deprecated-copy warning.
* | Eliminate inlining warning.Tor Egge2021-02-162-0/+4
|/
* Remove the sync interface from Bucket Executor. Due to the nature of ↵Henning Baldersheim2021-02-023-10/+5
| | | | | | requiring a bucket lock it is very hard to get sync to work in a safe way. Instead the users must do their own accounting as they know their own threading model.
* Add noexcept and some other minor code health.Henning Baldersheim2021-02-021-4/+4
|
* Revert "Properly track execution of BucketTasks and provide sync() and ↵Henning Baldersheim2021-02-021-4/+4
| | | | order… "
* Properly track execution of BucketTasks and provide sync() and orderly shutdown.Henning Baldersheim2021-02-021-4/+4
|
* Revert "Implement BucketExecutor::sync."Henning Baldersheim2021-02-021-4/+4
|
* Implement BucketExecutor::sync.Henning Baldersheim2021-02-011-4/+4
|
* Wire reporting of attribute resource usage all the way to the cluster ↵Geir Storli2021-01-291-0/+1
| | | | controller via the host info API.
* Restore lost comment.Tor Egge2021-01-221-0/+1
|
* Merge branch 'master' into toregge/port-to-gcc-8Henning Baldersheim2021-01-221-8/+7
|\
| * Merge pull request #16164 from vespa-engine/balder/split-testHenning Baldersheim2021-01-221-7/+7
| |\ | | | | | | - Splitt common test code.
| | * - Splitt common test code.Henning Baldersheim2021-01-221-7/+7
| | | | | | | | | | | | | | | - Splitt out HandlerTest. - Make JobTest parameterized to run with current way of using frozen bucket concept, or future using BucketExecutor and content layer bucket locking.
* | | contains member function on unordered set is not available when usingTor Egge2021-01-221-2/+4
|/ / | | | | | | gcc 8.
* / Track attribute resource usage.Tor Egge2021-01-225-7/+89
|/
* Add sync in destructor.Henning Baldersheim2021-01-211-1/+3
|
* Add DummyBucketExecutor for use in testing.Henning Baldersheim2021-01-213-0/+78
|
* Merge pull request #16112 from ↵Henning Baldersheim2021-01-205-1/+33
|\ | | | | | | | | vespa-engine/toregge/add-service-layer-host-info-reporter Add ServiceLayerHostInfoReporter.
| * Add operator== and external operator<< to storage::spi::ResourceUsage.Tor Egge2021-01-203-0/+25
| |
| * Add ServiceLayerHostInfoReporter.Tor Egge2021-01-202-1/+8
| |
* | Merge pull request #16098 from ↵Henning Baldersheim2021-01-206-10/+67
|\ \ | |/ |/| | | | | vespa-engine/balder/implement-bucketexecutor-for-filestormanager Implement bucketexecutor interface and.
| * Add debug dumping and other minor followup on PR comments.Henning Baldersheim2021-01-201-4/+4
| |
| * Add class comments.Henning Baldersheim2021-01-201-0/+3
| |
| * Implement bucketexecutor interface and.Henning Baldersheim2021-01-196-10/+64
| |
* | Fix forward declarations.Tor Egge2021-01-191-1/+1
|/
* Include exception where needed and use std:.make_uniqueHenning Baldersheim2021-01-187-46/+12
|
* GC ghost file that clion generated.Henning Baldersheim2021-01-181-6/+0
|
* Add failure detection and a sync method.Henning Baldersheim2021-01-153-6/+5
|
* Use forward declarations to reduce amount of included code.Henning Baldersheim2021-01-153-4/+11
|
* - Add an BucketExecutor interface taking a bucket and a BucketTask for ↵Henning Baldersheim2021-01-156-29/+48
| | | | | | | | | executing a task by the content layer. - Add method to register this interface. - Minor followups after PR review. - Avoid including the world when not necessary to avoid having to recompile everyting for the smallest unrelated change.
* Listener is deregistered when returned object is destroyed.Tor Egge2021-01-151-0/+1
|
* Add interface for resource usage listener to spi.Tor Egge2021-01-159-0/+150
| | | | Propagate resource usage to service layer.
* Remove unused revert method.Tor Egge2021-01-132-55/+0
|
* Remove loadmetrics indirection.Henning Baldersheim2020-11-261-1/+0
|
* Remove load type from ContextHenning Baldersheim2020-11-243-68/+59
|
* Steal the traces explicit and force moving of traces. Also hide access to ↵Henning Baldersheim2020-11-231-0/+1
| | | | the root.
* Enforce minimum bucked used bits at document metastore load timeTor Brede Vekterli2020-11-121-0/+17
|
* Greatly simplify bucket DB persistence provider bootstrap procedureTor Brede Vekterli2020-10-162-0/+19
| | | | | | | | | | | | | | | | | | | | | | The legacy bucket DB initialization logic was designed for the case where bucket information was spread across potentially millions of files residing on spinning rust drives. It was therefore async and running in parallel with client operations, adding much complexity in order to deal with a myriad of concurrency edge cases. Replace this with a very simple, synchronous init method that expects the provider to have the required information readily and cheaply available. This effectively removes the concept of a node's "initializing" state, moving directly from reported state Down to Up. Even though a node still technically starts up in Initializing state, we never end up reporting this to the Cluster Controller as the DB init completes before the RPC server stack is set up. Legacy bucket DB initializer code will be removed in a separate pass. Also simplify bucket DB interface contract for mutating iteration, indicating that it is done in an unspecified order.
* Remove partitions from SPI.Tor Egge2020-10-1415-228/+49
|
* Use std::mutex and std:.condition_variable and GC some unused code.Henning Baldersheim2020-10-132-15/+17
|
* Remove broken copy and move constructor and assignment operatoos on ↵Henning Baldersheim2020-10-071-0/+1
| | | | | | vespalib::Lock and vespalib::Monitor. Also repair broken usages of the same.
* Add noexcept as indicated by -WnoeexceptHenning Baldersheim2020-10-071-1/+1
|