From d77bcd1bd9b61489709b42fe09c6034bdc4e9660 Mon Sep 17 00:00:00 2001 From: Alvaro Vega Date: Thu, 19 Oct 2023 14:29:53 +0200 Subject: [PATCH 1/2] Update db_indexes.md --- doc/manuals/db_indexes.md | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/doc/manuals/db_indexes.md b/doc/manuals/db_indexes.md index 38563bd..7c4251e 100644 --- a/doc/manuals/db_indexes.md +++ b/doc/manuals/db_indexes.md @@ -16,6 +16,8 @@ It is recommended to create the following index in this collection: The performance difference can be dramatic for large sets of data. For instance, for a collection with around ~3000000 query execution time can drop from 3 seconds to 1 millisecond. +Since version 3.0.0 of cygnus an index named `cyg_raw_opt` is created in this way but depending on datamodel. + ## Indexes in the aggregated data collection It is recommended to create the following index in this collection: @@ -26,3 +28,5 @@ It is recommended to create the following index in this collection: The performance difference can be dramatic for large sets of data. For instance, for a collection with around ~6000000 query execution time can drop from 28 seconds to 20 millisecond. + +Since version 3.0.0 of cygnus index named `cyg_agg_opt` is created in this way but depending on datamodel. From 9a5886472dbb5d36a1adacc12d845a86216d874b Mon Sep 17 00:00:00 2001 From: Alvaro Vega Date: Thu, 19 Oct 2023 15:30:50 +0200 Subject: [PATCH 2/2] update doc with table --- doc/manuals/db_indexes.md | 17 +++++++++++++++-- 1 file changed, 15 insertions(+), 2 deletions(-) diff --git a/doc/manuals/db_indexes.md b/doc/manuals/db_indexes.md index 7c4251e..3e0c1ee 100644 --- a/doc/manuals/db_indexes.md +++ b/doc/manuals/db_indexes.md @@ -16,7 +16,14 @@ It is recommended to create the following index in this collection: The performance difference can be dramatic for large sets of data. For instance, for a collection with around ~3000000 query execution time can drop from 3 seconds to 1 millisecond. -Since version 3.0.0 of cygnus an index named `cyg_raw_opt` is created in this way but depending on datamodel. +Since version 3.0.0 of cygnus an index named `cyg_raw_opt` is created in this way but depending on datamodel: + +| datamodel | keys | +| :-------------- | :--------------------------------------------------------------- | +| DMBYSERVICEPATH | recvTime, entityId, entityType, attrName, attrType, attrValue | +| DMBYENTITY | recvTime, attrName, attrType, attrValue | +| DMBYATTRIBUTE | recvTime, attrType, attrValue | + ## Indexes in the aggregated data collection @@ -29,4 +36,10 @@ It is recommended to create the following index in this collection: The performance difference can be dramatic for large sets of data. For instance, for a collection with around ~6000000 query execution time can drop from 28 seconds to 20 millisecond. -Since version 3.0.0 of cygnus index named `cyg_agg_opt` is created in this way but depending on datamodel. +Since version 3.0.0 of cygnus index named `cyg_agg_opt` is created in this way but depending on datamodel: + +| datamodel | keys | +| :--------------- | :------------------------------------------------- | +| DMBYSERVICEPATH | entityId, entityType, attrName, resolution, origin | +| DMBYENTITY | attrName, resolution, origin | +| DMBYATTRIBUTE | resolution, origin |