aboutsummaryrefslogtreecommitdiffstats
path: root/searchcore
Commit message (Collapse)AuthorAgeFilesLines
* Change AttributeContext to use makeReadGuard method on attribute vectors.Tor Egge2018-04-208-18/+26
|
* Change target attribute type in ImportedAttributeVector toTor Egge2018-04-201-1/+4
| | | | ReadableAttributeVector.
* Use forward declared ReferenceAttribute.Tor Egge2018-04-202-2/+3
|
* Add ReadableAttributeVector interface used to create a short-lived read ↵Geir Storli2018-04-192-3/+6
| | | | | | guard over an attribute vector. Let AttributeVector and ImportedAttributeVector implement this interface.
* Change ImportedAttributeVector to no longer implement IAttributeVector API.Geir Storli2018-04-191-1/+1
| | | | All read access happens through short lived ImportedAttributeVectorReadGuard.
* Remove unused dependency.Henning Baldersheim2018-04-161-2/+0
|
* allow proton to drop sort data when sending query resultsHåvard Pettersen2018-04-114-18/+30
|
* allow not unpacking the whitelist iteratorHåvard Pettersen2018-04-101-0/+2
|
* Merge pull request #5515 from ↵Tor Egge2018-04-095-31/+116
|\ | | | | | | | | vespa-engine/geirst/ensure-that-only-a-single-prepare-restart-happens-at-the-same-time-in-proton Ensure that only a single prepare restart happens at the same time in…
| * Ensure that only a single prepare restart happens at the same time in proton.Geir Storli2018-04-095-31/+116
| |
* | Pass a safe docIdLimit to bitvector iterator for whitelist.Tor Egge2018-04-073-7/+8
| |
* | Don't estimate 0 hits in whitelist.Tor Egge2018-04-061-1/+1
| |
* | Adjust comment for setWhiteListBlueprint method.Tor Egge2018-04-061-1/+1
| |
* | Use active lids bitvector for whitelisting.Tor Egge2018-04-0616-166/+77
|/ | | | This replaces nested small numeric attribute vector for blaclisting.
* Merge pull request #5475 from ↵Tor Egge2018-04-068-20/+16
|\ | | | | | | | | vespa-engine/toregge/remove-nested-doctyperepo-up-typedef Remove DocumentTypeRepo::UP typedef.
| * Remove DocumentTypeRepo::UP typedef.Tor Egge2018-04-058-20/+16
| | | | | | | | Remove unneeded includes.
* | Merge pull request #5468 from ↵Geir Storli2018-04-064-9/+149
|\ \ | |/ |/| | | | | vespa-engine/toregge/add-extra-check-for-update-operations-when-document-type-changed Add extra check for update operations when document types can be different
| * Add comment explaining why we check identity instead of equality.Tor Egge2018-04-061-0/+5
| |
| * Extend unit test coverage to check validation of updates with fallback toTor Egge2018-04-051-11/+42
| | | | | | | | creating document.
| * Add extra check if document type for update operation might be differentTor Egge2018-04-054-9/+113
| | | | | | | | (e.g. too old or too new) to ensure that we are able to handle replay.
* | Reduce default value of indexing.semiunboundtasklimit.Tor Egge2018-04-051-2/+2
|/
* Use forward declared DocumentTypeRepo.Tor Egge2018-04-0429-16/+57
|
* Start using DocumentTypeRepoFactory.Tor Egge2018-04-042-2/+6
|
* Stop using nested typedef for shared pointer to const DocumentTypeRepo.Tor Egge2018-04-0445-93/+93
|
* Use shared pointer to const DocumentTypeRepo.Tor Egge2018-04-046-9/+11
|
* Nested namespace and default destructorsHenning Baldersheim2018-04-039-46/+21
|
* Revert "Hardwire matchengine to always be online."Geir Storli2018-03-212-3/+4
|
* Hardwire matchengine to always be online.Tor Egge2018-03-212-4/+3
|
* Remove unused methods for turning off matchengine.Tor Egge2018-03-207-117/+3
|
* Remove dead code.Tor Egge2018-03-141-1/+0
|
* Match engine should be up if cluster state for any bucketspace saysTor Egge2018-03-146-9/+26
| | | | that node is up.
* Add metrics for executor threading service.Tor Egge2018-03-066-2/+68
|
* Move ExecutorThreadingServiceStats to metrics directory.Tor Egge2018-03-066-3/+3
|
* Add class comment.Tor Egge2018-03-051-2/+7
| | | | Reduce indirections in using statement.
* Collect executor stats once per metrics update and pass the info to bothTor Egge2018-03-057-9/+90
| | | | | metrics update method for document db legacy metrics and to metrics update method for tagged document db metrics.
* Consider replay more expensive when preparing for restart.Tor Egge2018-03-011-3/+3
|
* Don't use semi-unbound task limit for summary executor when visibilityTor Egge2018-02-283-14/+6
| | | | delay is nonzero.
* Use blocking thread stack executor for document summary writing.Tor Egge2018-02-272-2/+3
|
* log performance events in separate threadHåvard Pettersen2018-02-236-16/+60
|
* Add --json option to vespa-gen-testdocs.Tor Egge2018-02-161-84/+203
|
* Remove unused functionality from vespa-gen-testdocs app.Tor Egge2018-02-151-203/+0
|
* Merge pull request #4989 from vespa-engine/balder/filter-qflagsHenning Baldersheim2018-02-091-1/+1
|\ | | | | Ensure that we filter and QFLAGS and do not automatically say we have…
| * Ensure that we filter and QFLAGS and do not automatically say we have the ↵Henning Baldersheim2018-02-091-1/+1
| | | | | | | | same capabilities as our master.
* | Compute nodes queried/replied more correctly during merge to get it more ↵Henning Baldersheim2018-02-091-1/+6
|/ | | | correct in MLD cases.
* Consider the whole space and let naming reflect it is an estimate.Henning Baldersheim2018-02-074-17/+18
|
* Ill placed brace.Henning Baldersheim2018-02-071-1/+2
|
* Merge pull request #4949 from vespa-engine/balder/cleanup-softtimeoutJon Bratseth2018-02-0712-71/+90
|\ | | | | Balder/cleanup softtimeout
| * Also keep track of nodes queried/replied.Henning Baldersheim2018-02-065-6/+14
| |
| * Prevent underflow due to returning endidHenning Baldersheim2018-02-061-1/+1
| |
| * Actually set numCovered docs.Henning Baldersheim2018-02-062-3/+3
| |