Perfomatic:API
From MozillaWiki
Contents
Overview
Staging server up at: http://graphs-stage2.mozilla.org
The graph server API is a REST API with a JSON-only response format.
The API is scheduled to launch in Q1.
Test server will be ready early Q1.
Using getdata.cgi?
If you are, things are going to change.
The biggest change is how we store tests. Currently, a test id can give you test results in time:value intervals. In the new architecture a test id will only get you a description of that test, nothing more.
In order to get test data, you have to know the test AND build AND machine you want results for. The current architecture assumes this based on test id.
For example, let's assume you want test results for Tp3 tests. Querying graphs.mozilla.org/api/test gives you a list of tests by build and OS.
Documented endpoints
List tests
- http://graphs.mozilla.org/api/test
- Returns list of all tests run per os per branch per machine.
- Example response:
{ 'stat':'ok', 'tests': [ { 'id':123, 'name':'Tp3', 'branch': { 'name':'1.9', 'id':3 }, 'os': { 'name':'WinXP', 'id':2 }, 'machine': { 'name':'qm-mini-ubuntu01 ', 'id':2 } }, { 'id':23, 'name':'Ts', 'branch': { 'name':'1.9', 'id':3 } 'os': { 'name':'Vista', 'id':3 } 'machine': { 'name':'qm-mini-xp01', 'id':2 } } ] }
Get test
- http://graphs.mozilla.org/api/test?testId=TESTID&branchid=BRANCHID&machine=MACHINEID
- Returns information for a specific testId and branch and machine
- Example response:
{ 'id':23, 'name':'Ts', 'branch': { 'name':'1.9', 'id':3 }, 'os': { 'name':'Vista', 'id':3 }, 'machine': { 'name':'qm-mini-xp01', 'id':2 } }
Get test runs (data)
- http://graphs.mozilla.org/api/test/runs?id=TESTID&branchid=BRANCHID&machineid=MACHINEID
- returns test runs for a specific test, build and machine, default date limit is 4 weeks, sorted by date descending
- Data is sent in a compressed form due to the large amount of information
- Indexes:
- 0: testrunid
- 1: build
- 0: buildid
- 1: refbuildid
- 2: changeset
- 2: date
- 3: average
- 4: run_number
- 5: annotations (array)
- 0: note
- 1: bugid
- Example response:
{ 'stat':'ok', 'test_runs': [ [98, [23, '2007072704', '572a00c0e251'], 1208830929, 92.4, [['Something happened here', 988740]]], [99, [23, '2007072704', '572a00c0e251'], 1208834503, 87.3] ] }
Get values for latest test run for a test
- http://graphs.mozilla.org/api/test/runs/latest?id=TESTID&branchid=BRANCHID&machineid=MACHINEID - returns test run values for the latest run test for a specific test, branch and machine id
- Example response
{ 'stat':'ok', 'id':33, 'values':[ { 'value':22.3, 'page':'www.yahoo.com' }, { 'value':29.1, 'page':'www.myspace.com' }, { 'value':39.3, 'page':'www.facebook.com' } ] }
Get test run information
- http://graphs.mozilla.org/api/test/runs/info/?id=TESTRUNID - returns test run information for a specific testRunId
- Example response
{ 'stat':'ok', 'testrun': { 'id':98, 'build': { 'id':23, 'build_id':'2007072704', 'changeset':'572a00c0e251' }, 'date_run':1208830929, 'average':92.4, 'annotations': [ { 'note':'Something happened here', 'bug_id':988740 } ] } }
Test run values (formerly discrete data)
- http://graphs.mozilla.org/api/test/runs/values?id=TESTRUNID
- returns test run results for a specific testRunId
- Example response
{ 'stat':'ok', 'id':33, 'values':[ { 'value':22.3, 'page':'www.yahoo.com' }, { 'value':29.1, 'page':'www.myspace.com' }, { 'value':39.3, 'page':'www.facebook.com' } ] }
Get test data by revision
- http://graphs.mozilla.org/api/test/runs/revisions?revision=10d2046d2b64&revision=d16525937c8b
- Example response
{ "stat": "ok", "revisions": { "d16525937c8b": { "tscroll": { "test_runs": { "MacOSX 10.5.8": [ [ 3794718, # test run id 20100421051331, # build id 1271853540, # timestamp 19719.5 # average value ] ], "WINNT 6.1": [ [ 3794854, 20100421053747, 1271856720, 11871.4 ] ], "Fedora 12 x64 - Constantine": [ [ 3794354, 20100421051525, 1271852700, 10063.200000000001 ] ], "WINNT 5.1": [ [ 3794853, 20100421053747, 1271856720, 13063.5 ] ] }, "name": "tscroll", "id": 71 }, }, } }
Error codes
- Default error responses are as follows:
{ 'stat':'fail', 'code':100, 'message':'Test does not exist' }