Commit message (Collapse) | Author | Age | Files | Lines | |
---|---|---|---|---|---|
* | Add parsing of ONNX Runtime session options to services.xml | Lester Solbakken | 2021-09-30 | 1 | -7/+10 |
| | |||||
* | Enable grouped distribution for vespa-feed-bm and vespa-redistribute-bm. | Tor Egge | 2021-09-29 | 7 | -39/+115 |
| | |||||
* | Remove estimates for moved documents during document redistribution. | Tor Egge | 2021-09-29 | 9 | -527/+2 |
| | |||||
* | Count moved documents in vespa-redistribute-bm. | Tor Egge | 2021-09-28 | 8 | -8/+247 |
| | |||||
* | Merge pull request #19306 from ↵ | Geir Storli | 2021-09-27 | 1 | -4/+4 |
|\ | | | | | | | | | vespa-engine/toregge/widen-fields-to-handle-larger-numbers-when-reporting-node-stats-in-bm-cluster Widen fields to handle larger numbers when reporting node stats in bm cluster. | ||||
| * | Widen fields to handle larger numbers when reporting node stats in bm cluster. | Tor Egge | 2021-09-27 | 1 | -4/+4 |
| | | |||||
* | | Track default content node bucket db stripe bits setting. | Tor Egge | 2021-09-27 | 1 | -1/+1 |
|/ | |||||
* | Merge pull request #19258 from ↵ | Henning Baldersheim | 2021-09-27 | 9 | -75/+493 |
|\ | | | | | | | | | vespa-engine/toregge/add-estimate-moved-docs-ratio-test Add estimate moved docs ratio unit test. | ||||
| * | Use static helper methods to instantiate CalculateMovedDocsRatio. | Tor Egge | 2021-09-23 | 4 | -23/+53 |
| | | |||||
| * | Add estimate moved docs ratio unit test. | Tor Egge | 2021-09-23 | 9 | -75/+463 |
| | | |||||
* | | Anonymous resident memory used by current process is only sampled on linux. | Tor Egge | 2021-09-27 | 1 | -0/+6 |
|/ | |||||
* | Remove explicit tracking of enum store and multi-value address space usage. | Geir Storli | 2021-09-21 | 7 | -49/+14 |
| | | | | | This has been replaced by generic tracking of address space usage among components in attributes vectors in all sub databases. | ||||
* | Rename functions to more consistent names. | Tor Egge | 2021-09-20 | 1 | -7/+6 |
| | | | | Adjust calculation of moved docs ratio for replace mode. | ||||
* | Add vespa-redistribute-bm, a single-process benchmark of document ↵ | Tor Egge | 2021-09-20 | 13 | -46/+769 |
| | | | | redistribution. | ||||
* | Add class comment for BmNodeStatsReporter. | Tor Egge | 2021-09-17 | 1 | -0/+4 |
| | |||||
* | Add BmNodeStats, containing info about documents on nodes, buckets | Tor Egge | 2021-09-17 | 14 | -2/+513 |
| | | | | and bucktets with pending ideal state operations. | ||||
* | Increase lifetime of BmClusterController. | Tor Egge | 2021-09-16 | 8 | -36/+58 |
| | |||||
* | Refactor BmFeeder to reduce code duplication. | Tor Egge | 2021-09-16 | 6 | -182/+106 |
| | |||||
* | Move portions of BMParams to BmFeedParams. | Tor Egge | 2021-09-15 | 10 | -264/+420 |
| | | | | Move more feed functions to BmFeeder and BmFeed. | ||||
* | Create persistence engine before creating document db in BmNode. | Tor Egge | 2021-09-14 | 1 | -1/+1 |
| | | | | This order is needed to pass a proper bucket executor to the document db. | ||||
* | Merge pull request #19118 from ↵ | Geir Storli | 2021-09-14 | 31 | -444/+698 |
|\ | | | | | | | | | vespa-engine/toregge/move-bm-feed-handler-ownership-from-bm-node-to-bm-cluster Move ownership of feed handlers from BmNode to BmCluster. | ||||
| * | Move shared code to base class. | Tor Egge | 2021-09-14 | 9 | -233/+137 |
| | | |||||
| * | Fix typos in comments. | Tor Egge | 2021-09-14 | 2 | -2/+2 |
| | | |||||
| * | Move ownership of feed handlers from BmNode to BmCluster. | Tor Egge | 2021-09-14 | 29 | -249/+599 |
| | | |||||
* | | Refactor to represent files with a generic header in a new class. | Geir Storli | 2021-09-14 | 1 | -25/+15 |
|/ | | | | Also move DIRECT I/O alignment setting to a common place. | ||||
* | Use uint32_t for node index and number of nodes. | Tor Egge | 2021-09-13 | 7 | -22/+25 |
| | |||||
* | Enable multiple benchmarks nodes in benchmark cluster. | Tor Egge | 2021-09-11 | 10 | -70/+95 |
| | | | | Only show first 10 errors in message bus. | ||||
* | Move ownership of BmNode to BmCluster. | Tor Egge | 2021-09-10 | 7 | -123/+230 |
| | |||||
* | Update copyright notices. | Tor Egge | 2021-09-10 | 11 | -11/+11 |
| | |||||
* | Refactor port number assignment. | Tor Egge | 2021-09-10 | 5 | -41/+98 |
| | |||||
* | Start moving portions of vespa-feed-bm app to searchcore_bmcluster library. | Tor Egge | 2021-09-10 | 42 | -1036/+1556 |
| | |||||
* | add more information to exception message | Arne Juul | 2021-08-30 | 1 | -1/+2 |
| | |||||
* | signal timeout instead of continuing triggering core dumps | Arne Juul | 2021-08-30 | 1 | -0/+4 |
| | |||||
* | Merge pull request #18859 from ↵ | Geir Storli | 2021-08-25 | 2 | -0/+11 |
|\ | | | | | | | | | vespa-engine/toregge/get-paged-setting-from-live-attribute-config Get paged setting from live attribute config. | ||||
| * | Get paged setting from live attribute config. | Tor Egge | 2021-08-25 | 2 | -0/+11 |
| | | |||||
* | | info -> debug | Henning Baldersheim | 2021-08-25 | 1 | -1/+1 |
| | | |||||
* | | Add some debug logging | Henning Baldersheim | 2021-08-25 | 1 | -1/+1 |
| | | |||||
* | | Add more information to error message | Henning Baldersheim | 2021-08-25 | 1 | -1/+1 |
|/ | |||||
* | Report max attribute address space usage in state explorer. | Geir Storli | 2021-08-23 | 5 | -12/+34 |
| | |||||
* | Expose address space usage for all attribute components in explorer. | Geir Storli | 2021-08-20 | 1 | -2/+3 |
| | |||||
* | Report max address space used in attribute vector components from content ↵ | Geir Storli | 2021-08-20 | 3 | -47/+39 |
| | | | | | | | nodes (proton) to the cluster controller. This is more generic than explicit address space values for enum store and multi value. This is used in the cluster controller to determine whether to block external feed. | ||||
* | Use max address space used in attribute vector components to determine ↵ | Geir Storli | 2021-08-20 | 10 | -111/+45 |
| | | | | whether to block put and update operations in proton. | ||||
* | Add metric for the max address space used among components in all attribute ↵ | Geir Storli | 2021-08-20 | 3 | -4/+9 |
| | | | | vectors in a document db. | ||||
* | Track max address space usage among components in attributes vectors in all ↵ | Geir Storli | 2021-08-20 | 7 | -9/+86 |
| | | | | sub databases. | ||||
* | Also provide used and on-hold memory in overview. | Henning Baldersheim | 2021-08-15 | 1 | -0/+2 |
| | |||||
* | Wire in the shared executor for use when loading attributes. | Henning Baldersheim | 2021-08-13 | 4 | -11/+18 |
| | |||||
* | Add an executor to the AttributeVector::load/onLoad interface so attributes ↵ | Henning Baldersheim | 2021-08-12 | 2 | -2/+2 |
| | | | | can use multithread load if feasible. | ||||
* | Add some debug logging. | Henning Baldersheim | 2021-08-09 | 3 | -2/+15 |
| | |||||
* | Split current global_filter_limit into global_filter.lower_limit/upper_limit. | Henning Baldersheim | 2021-08-04 | 3 | -12/+18 |
| | | | | | | | | | If estimated_hits < lower_limit no filter is set which will cause fallback to bruteforce. If estimated_hits in [lower_limit, upper_limit] apply global filter. if estimated_hits > upper_limit an empty filter is set. This will avoid the filter setup cost. So if the filter has a huge setup cost, you can reduce upper_limit to a number below 1.0 and instead increase target_num_hits similarly. Setting target_num_hits to 1.0/upper_limit * 1.2 should give similar recall. This will add a 20% safety to handle correlation of filter and NearestNeightbor calculation. | ||||
* | Also perform topk when you do not use a global filter. | Henning Baldersheim | 2021-08-03 | 1 | -0/+4 |
| |