Jump to: navigation, search

FLOP:CVE Monitoring

No change in size, 1 year ago
no edit summary
The <tt>cver</tt> tool is currently stateless: it takes some bytes and it makes some bytes. Of course, <tt>jira</tt> and <tt>mongoDB</tt> are not, and their states must be kept in sync. Does <tt>cver</tt> require its own set of <tt>mongoDB</tt> collections to maintain the sync? This probably the most challenging aspect of the proposal.
- * every update of the <tt>cve-search</tt> database must trigger an update of <tt>jira</tt>- * every CRUD path of <tt>cve-search</tt> must have an equivalent CRUD path of <tt>jira</tt>- * the sync of <tt>cve-search</tt> and <tt>jira</tt> must be always provable
However, we don't need to deal with <tt>cve-search</tt> directly: we can transform it into an intermediate state associated with <tt>cver</ver> that has its own paths, and then make the equivalent <tt>jira</tt> path from those. We may just 'bulk transform' the <tt>cve-search</tt> to a (probably much simpler) schema more directly related to that of a <tt>jira</tt> issue. We just need a collection of what are essentially <tt>jira</tt> records, with meta data to control the sync.

Navigation menu