add algorithm test bench (vesoft-inc#39)

* add algorithm test bench * refactor yml name
cjdxlgb · Jan 30, 2022 · e87f9d8 · e87f9d8
1 parent fa91a9f
commit e87f9d8
Show file tree

Hide file tree

Showing 4 changed files with 55 additions and 0 deletions.
diff --git a/.github/workflows/maven.yml → .github/workflows/pull_request.yml b/.github/workflows/maven.yml → .github/workflows/pull_request.yml
diff --git a/.github/workflows/deploy_release.yml → .github/workflows/release.yml b/.github/workflows/deploy_release.yml → .github/workflows/release.yml
diff --git a/.github/workflows/deploy_snapshot.yml → .github/workflows/snapshot.yml b/.github/workflows/deploy_snapshot.yml → .github/workflows/snapshot.yml
diff --git a/doc/algorithm_test.md b/doc/algorithm_test.md
@@ -0,0 +1,55 @@
+# Nebula Algorithm test
+
+## Data preparation
+
+1. Install Nebula Graph 3.0.0
+    * cluster info:
+
+    `3 metad, 3 graphd, 3 storaged`
+
+    * space info:
+
+    |Name	    |Partition Number|Replica Factor|Charset|Vid Type|
+    |-----------|----------------|--------------|-------|--------|
+    |algo_space	|       100	     |     1	    |  uft8	|  int64 |
+
+2. Put LDBC sf100 dataset to HDFS: /ldbc/sf100
+
+    We use Nebula Exchange to import ldbc sf100 dataset into Nebula, importing details see: https://github.com/vesoft-inc/nebula-exchange/tree/master/bench
+
+
+## Test algorithms with HDFS datasource and HDFS data sink
+We use LDBC's `forum_hasMember_person` edge dataset to test the algorithms.
+
+ |     algorithms   |   data amount  | concurrency |executor-memory|iter|    other params      | duration|
+ |------------------|----------------|-------------|---------------|----|----------------------|---------|
+ |  pagerank	    | 180 millions	 |    30 	   |     45G       | 10 |resetProb:0.15        | 1.1min  |
+ |  trianglecount   | 180 millions   |    30 	   |     45G       | -  |      -               |  19min  |
+ |       lpa        | 180 millions	 |    30 	   |     45G       | 10 |      -               |  11min  |
+ |       wcc        | 180 millions	 |    30 	   |     45G       | 10 |      -               |   40s   |
+ |       scc   	    | 180 millions	 |    30 	   |     45G       | 10 |      -               |   38s   |
+ |clustercoefficient| 180 millions	 |    30 	   |     45G       | -  |      -               |  25min  |
+ |      kcore	    | 180 millions	 |    30 	   |     45G       | 10 |   degree:1           |   1min  |
+ | degreeCentrality | 180 millions	 |    30 	   |     45G       | -  |      -               |   34s   |
+ |    louvain	    | 180 millions	 |    30 	   |     45G       | 10 |internalIter:5,tol:0.5|  10min  |
+ |       BFS	    | 180 millions	 |    30 	   |     45G       | 10 |root:17592186046139   |   52s   |
+ |      HANP	    | 180 millions	 |    30 	   |     45G       | 10 |hop:0.1,preference:1.0|  17min  |
+ |     closeness    | 180 millions	 |    30 	   |     45G       | -  |      -               |  17min  |
+
+# Test algorithms with Nebula datasource and Nebula data sink
+We use Nebula's `HAS_MEMBER` edge type to test the algorithms, and the result is writen into new tag.
+
+ |     algorithms   |   data amount  | concurrency |executor-memory|iter|    other params      | duration|
+ |------------------|----------------|-------------|---------------|----|----------------------|---------|
+ |  pagerank	    | 180 millions	 |    30 	   |     45G       | 10 |resetProb:0.15        |  5.2min |
+ |  trianglecount   | 180 millions   |    30 	   |     45G       | -  |      -               |  12min  |
+ |       lpa        | 180 millions	 |    30 	   |     45G       | 10 |      -               |  44min  |
+ |       wcc        | 180 millions	 |    30 	   |     45G       | 10 |      -               |  1.8min |
+ |       scc   	    | 180 millions	 |    30 	   |     45G       | 10 |      -               |  2min   |
+ |clustercoefficient| 180 millions	 |    30 	   |     45G       | -  |      -               |  11min  |
+ |      kcore	    | 180 millions	 |    30 	   |     45G       | 10 |   degree:1           |  3.2min |
+ | degreeCentrality | 180 millions	 |    30 	   |     45G       | -  |      -               |  3.6min |
+ |    louvain	    | 180 millions	 |    30 	   |     45G       | 10 |internalIter:5,tol:0.5|  14min  |
+ |       BFS	    | 180 millions	 |    30 	   |     45G       | 10 |root:17592186046139   |  2.3min |
+ |      HANP	    | 180 millions	 |    30 	   |     45G       | 10 |hop:0.1,preference:1.0|  52min  |
+ |     closeness    | 180 millions	 |    30 	   |     45G       | -  |      -               |  42min  |