IntenseFinance/nosqlbench
3
0
Fork 0
mirror of https://github.com/nosqlbench/nosqlbench.git synced 2024-12-23 15:40:44 -06:00
Code Issues Packages Projects Releases Wiki Activity

disabling strictness with stable api, workload partial updates

Browse Source
This commit is contained in:
Madhavan Sridharan 2023-10-13 17:12:45 -04:00
parent 5667e0570a
commit 57e7eaab05
2 changed files with 46 additions and 16 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
adapter-mongodb/src/main/java/io/nosqlbench/adapter/mongodb/core/MongoSpace.java
View File

@ -87,8 +87,8 @@ public class MongoSpace implements NBNamedElement, AutoCloseable {
// https://www.mongodb.com/docs/v7.0/reference/stable-api
ServerApi serverApi = ServerApi.builder()
.version(ServerApiVersion.V1)
.deprecationErrors(true)
.strict(true)
.deprecationErrors(false)
.strict(false)//Needed because createSearchIndexes is not in stable API
.build();
MongoClientSettings settings = MongoClientSettings.builder()

58
adapter-mongodb/src/main/resources/activities/baselinesv2/mongodb_vector_search.yaml
View File

@ -11,9 +11,13 @@ description: |
scenarios:
default:
schema: run driver=mongodb tags==block:"schema.*" threads==1 cycles==UNDEF database=baselines
rampup: run driver=mongodb tags==block:rampup cycles===TEMPLATE(rampup-cycles,100000) threads=auto database=baselines
rampup: run driver=mongodb tags==block:rampup cycles===TEMPLATE(trainsize) threads=auto database=baselines errors=counter,retry
main: run driver=mongodb tags==block:'main-.*' cycles===TEMPLATE(main-cycles,100000) threads=auto database=baselines
drop: run driver=mongodb tags==block:drop-collection threads==1 cycles==UNDEF database=baselines
drop: run driver=mongodb tags==block:drop-entire-collection threads==1 cycles==UNDEF database=baselines
search_and_index: >-
run driver=mongodb alias=search_and_index tags='block:main-read' labels='target:mongodbatlas'
cycles=TEMPLATE(testsize) errors=counter,retry stride=100 striderate=7.50
read_ratio=1 threads=500 database=baselines
params:
instrument: true
@ -30,6 +34,7 @@ bindings:
train_floatlist: HdfFileToFloatList("testdata/TEMPLATE(dataset).hdf5", "/train"); ToCqlVector();
#READ
test_floatlist: HdfFileToFloatList("testdata/TEMPLATE(dataset).hdf5", "/test"); ToCqlVector();
relevant_indices: HdfFileToIntArray("testdata/TEMPLATE(dataset).hdf5", "/neighbors")
blocks:
schema:
@ -65,8 +70,8 @@ blocks:
}
}
},
validationLevel: "off",
validationAction: "warn",
validationLevel: "strict",
validationAction: "error",
comment: "keyvalue collection creation with strict types and a required 'key' field."
}
create_key_index: |
@ -121,26 +126,46 @@ blocks:
}
main-read:
params:
ratio: 5
ratio: TEMPLATE(read_ratio,5)
ops:
main-select: |
{
aggregate: "TEMPLATE(collection,keyvalue)",
pipeline: [
$vectorSearch: {
index: "kv_value_vector_search_idx",
path: "value",
queryVector: {test_floatlist},
numCandidates: TEMPLATE(num_candidates,1000)
limit: TEMPLATE(top_k,100)
"$vectorSearch": {
"index": "kv_value_vector_search_idx",
"path": "value",
"queryVector": {test_floatlist},
"numCandidates": TEMPLATE(num_candidates,1000)
"limit": TEMPLATE(top_k,100)
},
{
"$project": {
"_id": 0,
"key": 1,
"value": 1,
"score": { "$meta": "vectorSearchScore" }
}
}
],
readConcern: { level: "majority" },
readConcern: { "level": "majority" },
comment: "Find the results for the given 'value' vector search embedding."
}
verifier-init: |
k=TEMPLATE(top_k,100)
relevancy=scriptingmetrics.newRelevancyMeasures(_parsed_op);
relevancy.addFunction(io.nosqlbench.engine.extensions.computefunctions.RelevancyFunctions.recall("recall",k));
relevancy.addFunction(io.nosqlbench.engine.extensions.computefunctions.RelevancyFunctions.precision("precision",k));
relevancy.addFunction(io.nosqlbench.engine.extensions.computefunctions.RelevancyFunctions.F1("F1",k));
relevancy.addFunction(io.nosqlbench.engine.extensions.computefunctions.RelevancyFunctions.reciprocal_rank("RR",k));
relevancy.addFunction(io.nosqlbench.engine.extensions.computefunctions.RelevancyFunctions.average_precision("AP",k));
verifier: |
actual_indices=cql_utils.cqlStringColumnToIntArray("key",result);
relevancy.accept({relevant_indices},actual_indices);
return true;
main-write:
params:
ratio: 5
ratio: TEMPLATE(write_ratio,5)
ops:
main-insert: |
{
@ -154,8 +179,13 @@ blocks:
writeConcern: { w: "majority" },
comment: "Insert documents into keyvalue collection."
}
drop-collection:
drop-entire-collection:
ops:
drop-vsearch-index: |
{
dropSearchIndex: "TEMPLATE(collection,keyvalue)",
name: "kv_value_vector_search_idx"
}
drop-collection: |
{
drop: "TEMPLATE(collection,keyvalue)",