summaryrefslogtreecommitdiff
path: root/lib/custodian
AgeCommit message (Collapse)Author
2013-02-05Adding (currently broken) open smtp relay check for steve to look at.john hackett
2013-01-24 fixedSteve Kemp
2013-01-24 New releaseSteve Kemp
2013-01-24 Store the duration of tests.Steve Kemp
2013-01-24 Updated to log duration of test-runs with our alerters, if theySteve Kemp
implement a "duration" method.
2013-01-14 Remove the mauve-handlingSteve Kemp
2013-01-14 fixed comments.Steve Kemp
2012-12-31 Added redis protocol testSteve Kemp
2012-12-24 Added reference to missing URI librarySteve Kemp
2012-12-20 It is a bug for a test to have a protocol not matching theSteve Kemp
scheme of the URL. e.g. This is wrong: https://example.com/ must run http .. ("https" != "http").
2012-12-19 Ensure that our notification periods are integersSteve Kemp
2012-12-17 Ensure our test fails if the content doesn't matchSteve Kemp
2012-12-12 The suppression periods are now configurable tooSteve Kemp
2012-12-12 The suppression period varies between working and out of hoursSteve Kemp
2012-12-12 Updated comment.Steve Kemp
2012-12-12 The settings object will let you retrieve an arbitrary key.Steve Kemp
2012-12-12 The dequeuing process passes the settings to the wroker.Steve Kemp
2012-12-12 Allow the alerter to receive the settings object.Steve Kemp
2012-12-12 Use the SHA1 hash of the input line as the alert ID for mauve.Steve Kemp
2012-12-11 Updated to be more explicit in the error field.Steve Kemp
2012-12-11 Updated comments.Steve Kemp
2012-12-11 Updated to quote the expected results in the details field.Steve Kemp
2012-12-11 Use and require an exact match in DNS results.Steve Kemp
2012-12-10 Use the parser-input as the alert-idSteve Kemp
2012-12-10 Ensure we raise an exception if we're trying to resolve an unknown ↵Steve Kemp
record-type. Added test-case too.
2012-12-10Ensure DNS test returns nil if an unrecognised record type is given,Patrick J Cherry
to make sure error handling in the caller works.
2012-12-07 When reporting on failure-types the HTTP/HTTPS test is often reported ↵Steve Kemp
incorrectly. Fixed this now, based on the URL and added a test case.
2012-12-06 We don't set the .detail field unless we're raising an alert.Steve Kemp
This will mean historically we can see what the failure reason was.
2012-11-30 comment + neatness update.Steve Kemp
2012-11-30 Hard-wire repeat for the moment. It'll come from the configurationSteve Kemp
file shortly.
2012-11-29 * Added mauve suppression to the mauve alert class.Patrick J Cherry
* Added debian recommendation for mauvealert-client
2012-11-27 Expire keys sanely.Steve Kemp
2012-11-27 Ensure the telnet test is loaded.Steve Kemp
2012-11-27 Added telnet test.Steve Kemp
2012-11-26 Ensure we load the resolving librarySteve Kemp
2012-11-26 Updated so that each alert-type has its dedicated alert-target.Steve Kemp
2012-11-26 The target is used to specify the redis hostname - use it.Steve Kemp
2012-11-26 Don't attempt to raise/clear with redis/mauve unless they were loaded.Steve Kemp
2012-11-26 New releaseSteve Kemp
2012-11-26 Show failing lines.Steve Kemp
2012-11-26 Updated to make the requirement of redis soft, as it will probably beSteve Kemp
not used.
2012-11-26 Allow multiple alerters.Steve Kemp
2012-11-26 Ensure we can be loaded, even if there is no redis rubygem present.Steve Kemp
2012-11-26 Added the redis-alerter.Steve Kemp
2012-11-26 Explicitly watch a "tube" - a named queue - to avoid stomping onSteve Kemp
anything else running under beanstalkd.
2012-11-25 Use the user-supplied text, if present.Steve Kemp
2012-11-25 Parse the "otherwise 'xxxx'." part of the tests.Steve Kemp
2012-11-25 Removed references to sentinel.Steve Kemp
2012-11-24 Removed test/driver-code.Steve Kemp
2012-11-24 Implemented outgoing mail sending.Steve Kemp