Datastore Collections

CodeMetrics uses the following collections across its datastore. These are created automatically on first access unless DATASTORE_AUTO_CREATE=false is set (see Datastores).

Collection Name	Used By
ado-issues-by-date	Azure DevOps Issues
alerts	GitHub Dependency Alerts
commit-prs	VCS (all providers)
deploy-bounds	Deployments
dynatrace-events	Dynatrace Pipelines
earliest-commits	VCS (all providers)
fetch-file	VCS (all providers)
fetch-merge-rules	VCS (all providers)
github-issues	GitHub Issues
issues	Jira Issues, ServiceNow Issues
pipeline-executions	Azure Pipelines, GitHub Pipelines
pipelines-job-names	Pipelines (all providers)
queries	Stored Queries
repo-changes	Bitbucket Server VCS, GitHub VCS
repo-commits	Bitbucket Server VCS, GitHub VCS
token_ids	Authentication (long-lived token IDs)
tokens	Authentication (server tokens)
vcs-cache	VCS (all providers)
vulns	Vulnerabilities

Collection naming by implementation

The collection name shown in the table above is the logical name used in code. Each datastore implementation maps this to a physical resource as follows.

DynamoDB

Each collection becomes a separate DynamoDB table. The table name is formed by combining a configurable prefix with the collection name:

{DATABASE_NAME}_{collectionName}

The prefix defaults to CodeMetrics if DATABASE_NAME is not set. For example, the vcs-cache collection becomes the table CodeMetrics_vcs-cache.

MongoDB

Each collection is stored as a MongoDB collection within a single database. The collection name is used as-is; only the database name is configurable:

DATABASE_NAME  (default: code-metrics)
└── collectionName  (e.g. vcs-cache)

For example, the vcs-cache collection is stored in the code-metrics database as the collection vcs-cache.

NeDB (local file database)

Each collection is stored as a separate file on disk. The file name is formed from a fixed prefix and the collection name:

{DATASTORE_PATH}/code-metrics-{collectionName}.db

For example, the vcs-cache collection is stored at {DATASTORE_PATH}/code-metrics-vcs-cache.db. If DATASTORE_PATH is not set, NeDB operates in-memory.

In-memory

Each collection is held in process memory under its logical name with no transformation. No physical resource is created.