Commit message (Collapse) | Author | Age | Files | Lines | |
---|---|---|---|---|---|
* | - Complete dumping of 1 index field before progressing to the next. | Henning Baldersheim | 2024-02-08 | 2 | -16/+17 |
| | | | | | - This prevents allocating memory buffers, and file descriptors for all fields concurrently. - It will reduce memory footprint during flush if there are many fields. | ||||
* | Add low-level benchmark program for search iterators. | Geir Storli | 2024-02-08 | 2 | -7/+7 |
| | | | | | Currently, it can benchmark the following query operators over an attribute vector: Single term, In, WeightedSet, DotProduct, Or. | ||||
* | make default flow stats more explicit | Håvard Pettersen | 2024-02-06 | 1 | -0/+3 |
| | | | | | | for both simple and complex leafs account for number of inner children in complex leafs account for seek nesting for complex leafs with children | ||||
* | Merge pull request #29976 from ↵ | Arne H Juul | 2024-01-24 | 1 | -10/+59 |
|\ | | | | | | | | | vespa-engine/arnej/unit-test-verify-ranksetup-streaming write vsmfields.cfg and add smoke test | ||||
| * | unit test streaming mode where possible | Arne Juul | 2024-01-19 | 1 | -15/+22 |
| | | |||||
| * | write vsmfields.cfg and add smoke test | Arne Juul | 2024-01-19 | 1 | -6/+48 |
| | | |||||
* | | wire in strict flow analysis and strict-aware sorting | Håvard Pettersen | 2024-01-22 | 1 | -3/+4 |
|/ | | | | | | | | | | strict_cost added to all blueprints separate top-down sort step after optimize move relative estimate out of blueprint state optimize all children; to calculate flow stats leaf defaults: matching>0.9: est: 0.5, cost: 1.0, strict_cost: 1.0 matching<=0.9: est: rel_est, cost: 1.0, strict_cost: rel_est | ||||
* | Add feature flag for allow sorting blueprints by cost estimate instead of ↵ | Henning Baldersheim | 2023-12-19 | 1 | -232/+119 |
| | | | | est_hits. | ||||
* | Remove most of the now void clock indirection. | Henning Baldersheim | 2023-12-15 | 7 | -20/+18 |
| | |||||
* | Unify on using reference where possible. | Henning Baldersheim | 2023-12-12 | 1 | -3/+3 |
| | |||||
* | Wire in thread bundle to execute info and request context. | Henning Baldersheim | 2023-12-12 | 2 | -9/+12 |
| | |||||
* | Revert "Revert "relative estimate"" | Henning Baldersheim | 2023-12-11 | 1 | -4/+19 |
| | |||||
* | Revert "relative estimate" | Henning Baldersheim | 2023-12-09 | 1 | -19/+4 |
| | |||||
* | relative estimate | Håvard Pettersen | 2023-12-08 | 1 | -4/+19 |
| | |||||
* | Merge pull request #29551 from ↵ | Henning Baldersheim | 2023-12-05 | 2 | -2/+2 |
|\ | | | | | | | | | vespa-engine/balder/gc-use-shared-executor-for-warmup Use shared executor for warmup and GC warmup executor. | ||||
| * | Use shared executor for warmup and GC warmup executor. | Henning Baldersheim | 2023-12-05 | 2 | -2/+2 |
| | | |||||
* | | - Control creation of temporary postinglists during fetchPostings for ↵ | Henning Baldersheim | 2023-12-04 | 1 | -3/+3 |
|/ | | | | non-strict iterators. | ||||
* | Avoid timeout during grouping leaving distributionKey unset. Populate it ↵ | Henning Baldersheim | 2023-11-30 | 1 | -3/+3 |
| | | | | right after completing grouping. | ||||
* | Add InTerm to backend. | Tor Egge | 2023-11-24 | 2 | -0/+3 |
| | |||||
* | Avoid dereferencing first item in an empty vector. | Henning Baldersheim | 2023-11-20 | 1 | -1/+1 |
| | |||||
* | - We are now always nesting multivalue grouping for indexed search. | Henning Baldersheim | 2023-11-20 | 1 | -14/+15 |
| | |||||
* | Add flag for marking phrase always expensive. | Henning Baldersheim | 2023-11-19 | 1 | -1/+33 |
| | |||||
* | Merge pull request #29369 from vespa-engine/balder/gc-unused-split-parameter | Henning Baldersheim | 2023-11-17 | 1 | -24/+8 |
|\ | | | | | Fully GC unused parameter as we now always split phrases. | ||||
| * | Fully GC unused parameter as we now always split phrases. | Henning Baldersheim | 2023-11-17 | 1 | -24/+8 |
| | | |||||
* | | If hit_rate is below 1% drop match phase limiting. It has too high fixed ↵ | Henning Baldersheim | 2023-11-16 | 1 | -0/+1 |
|/ | | | | cost and will liklely make things worse. | ||||
* | Merge pull request #29284 from ↵ | Tor Brede Vekterli | 2023-11-09 | 1 | -10/+25 |
|\ | | | | | | | | | vespa-engine/vekterli/include-doctype-and-gid-with-metadata-doc-entries Include doc type name and GID in metadata iteration results | ||||
| * | Simplify by passing in and storing the `DocTypeName` verbatim | Tor Brede Vekterli | 2023-11-08 | 1 | -42/+4 |
| | | |||||
| * | Include doc type name and GID in metadata iteration results | Tor Brede Vekterli | 2023-11-08 | 1 | -9/+62 |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | Document type is fetched from the associated `IPersistenceHandler` on-demand; it is assumed the lifetime of the pointer must be valid for the entire lifetime of the iterator itself, as the latter holds a valid handler snapshot. For simplicity, it's possible to _not_ pass in a handler, in which case the doc type name will be implicitly empty. Some expected `DocEntry` sizes have been adjusted, as we now report the size of the document type and GID alongside the base type size. | ||||
* | | Sameelement behaves like an and with extra constraints. | Henning Baldersheim | 2023-11-07 | 2 | -6/+5 |
|/ | | | | So it should behave the sameway during fetchPostings too. | ||||
* | Merge pull request #29269 from ↵ | Geir Storli | 2023-11-07 | 5 | -47/+75 |
|\ | | | | | | | | | vespa-engine/geirst/control-resource-usage-when-in-maintenance Control resource usage when node in maintenance | ||||
| * | Also tune or turn off background jobs when content node is in maintenance. | Geir Storli | 2023-11-07 | 4 | -19/+56 |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Previously the following has been adjusted when the node is retired: 1) Lid space compaction - turned off. 2) Flush engine strategy - tuned to reduce disk and CPU usage. 3) Attribute vector compaction - tuned to reduce memory allocations and CPU usage. In a node retirement scenario documents are being removed from the node, and eventually the node is deleted. Without the adjustments above a lot of resources are spent "fixing" the results of removing documents, and the process just takes a lot longer. A similar set of challenges can occur when a node is set in maintenance, especially if the node transitions from retired to maintenance. E.g. this happens when the Vespa version is upgraded in Vespa Cloud. With this change the resource usage of background jobs are kept in check for both a retired node and a node in maintenance. | ||||
| * | Rewrite to use GTest. | Geir Storli | 2023-11-07 | 2 | -35/+26 |
| | | |||||
* | | Merge pull request #29266 from ↵ | Henning Baldersheim | 2023-11-07 | 2 | -4/+11 |
|\ \ | | | | | | | | | | | | | vespa-engine/vekterli/expose-doc-type-name-from-persistence-handler Expose document type name from `IPersistenceHandler` interface | ||||
| * | | Expose document type name from `IPersistenceHandler` interface | Tor Brede Vekterli | 2023-11-07 | 2 | -4/+11 |
| |/ | |||||
* | | Merge pull request #29264 from ↵ | Henning Baldersheim | 2023-11-07 | 1 | -0/+3 |
|\ \ | | | | | | | | | | | | | vespa-engine/toregge/extend-persistence-conformance-test-wrt-remove-by-gid Test remove by gid for nonexisting gid and for gid with tombstone. | ||||
| * | | Test remove by gid for nonexisting gid and for gid with tombstone. | Tor Egge | 2023-11-07 | 1 | -0/+3 |
| |/ | |||||
* / | If match-phase limiting has concluded that a post filter is most efficient, | Henning Baldersheim | 2023-11-07 | 1 | -2/+3 |
|/ | | | | we must only generate posting lists if it is actually benefiscal. If not the fixed cost is too high. | ||||
* | Add removeByGidAsync() to spi. | Tor Egge | 2023-11-06 | 2 | -0/+10 |
| | |||||
* | Make test for target numdocs auto-tuning aware of sanitizers | Tor Brede Vekterli | 2023-11-06 | 1 | -2/+14 |
| | | | | | Test must mirror the logic used for enabling auto-tuning, or it will fail when testing on a sanitizer-instrumented build. | ||||
* | - Add maxHits param that can be used to compute a realistic(low) hit rate ↵ | Henning Baldersheim | 2023-11-03 | 2 | -9/+11 |
| | | | | when fetching summaries (summary features). | ||||
* | Move `HwInfo` from `proton` namespace to `vespalib` | Tor Brede Vekterli | 2023-11-01 | 19 | -12/+29 |
| | | | | | This is information that is valuable to many different components, not just the search core internals. | ||||
* | Merge pull request #29156 from ↵ | Geir Storli | 2023-10-31 | 2 | -5/+18 |
|\ | | | | | | | | | vespa-engine/balder/test-document-refusing-to-move Test that a document that refuses to move is not moved. | ||||
| * | Test that a document that refuses to move is not moved. | Henning Baldersheim | 2023-10-30 | 2 | -5/+18 |
| | | |||||
* | | Get attribute vector match settings from live config when starting proton. | Tor Egge | 2023-10-30 | 1 | -4/+108 |
|/ | |||||
* | Test that documents failing move are detected and causes retry and eventual ↵ | Henning Baldersheim | 2023-10-30 | 3 | -6/+68 |
| | | | | completition. | ||||
* | BucketMove is 3 phased, startMove, createMoveOps, completeMove. | Henning Baldersheim | 2023-10-30 | 3 | -5/+7 |
| | | | | | | | First and last happens in master, while second happens in bucket executor. As feed might happen inbetween these phases, care must be taken to check that the state for a document is the same in complete as it was in first phase. If not the move must be retried. | ||||
* | Add test for proper deduction of target num docs. | Henning Baldersheim | 2023-10-27 | 1 | -34/+75 |
| | |||||
* | Allow longer timeout to allow tests to complete on a heavily loaded system. | Henning Baldersheim | 2023-10-18 | 1 | -2/+2 |
| | |||||
* | Consolidate extraction of tokens from annotated string field value. | Tor Egge | 2023-10-10 | 1 | -1/+1 |
| | |||||
* | Merge pull request #28840 from ↵ | Tor Egge | 2023-10-10 | 1 | -17/+213 |
|\ | | | | | | | | | vespa-engine/toregge/flush-indexes-when-enabling-interleaved-features Flush memory indexes to disk then fusion disk indexes as soon as |