mapbox / sumo Goto Github PK

w/ documentation and tests

I noticed some throttling messages sneaking into the console on a search with a very large number of results, along the lines of

> sumo --query '_sourceCategory = my-thing' --from 4d

{ [HTTPError: Response code 429 (Too Many Requests)]
  message: 'Response code 429 (Too Many Requests)',
  host: 'api.sumologic.com',
  hostname: 'api.sumologic.com',
  method: 'GET',
  path: '/api/v1/search/jobs/4FD8AA773A30EAD6',
  statusCode: 429,
  statusMessage: 'Too Many Requests' }

The search continued to return results.

Trying a CLI query like

 sumo -q '_sourceCategory = my-thing | order by _messageTime asc' --from 4d

... pagination doesn't seem to work properly. It appears that each step back in time returns messages up until present, so it keeps repeat printing the most recent messages.

I am seeing that the code in my proof of concept using this library gets this error from node:

Unhandled promise rejection (rejection id: 2): Error: No records were found after 30s

I can't see a way to handle this rejection as a consumer of this library.
I think mapbox/sumo should handle all internal errors and propagate any relevant errors. The relevant errors should be propagated as a promise rejection for sumo.search calls and an error event for sumo.createReadStream