| Age | Commit message (Collapse) | Author |
|---|
|
|
|
This ensures that all tests always run, and we have an ordering.
|
|
|
|
This means that tests will only ever be enqueued once, regardless
of how many times they are parsed.
In the past we could have a configuration file that read:
test1 ..
test2 ..
test3 ..
Parsing/adding this file would result in a queue looking like so:
test1 ..
test2 ..
test3 ..
test1 ..
test2 ..
test3 ..
test1 ..
test2 ..
test3 ..
Now the queue will *ALWAYS* look like this:
test1 ..
test2 ..
test3 ..
In the normal course of events this won't matter, as teh processing
loop will look like so:
* Add new jobs every minute.
* Worker runs the jobs.
In the case of a failing job though the test might take 2.5 minutes
and that will cause the queue to backup. (2.5 minutes because a test
is repeated 5 times before a fail is announced, and the timeout is
30 seconds. These values can and should be tweaked.)
With the new method even if the queue is slowly draining the queue
will never grow to containu hundreds of events it will just be "topped
up" not "overflowing".
Thanks to James Hannah for the suggestion, and James Lawrie for
the patience.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
This will make visualization more simple.
|
|
This allows our configuration file `/etc/custodian/custodian.cfg`
to contain something like this, without errors;
alerter = file , redis
|
|
This is a good thing to do, as Chrome will apaprently be
refusing to show sites with SHA-1 in use over SHA-256.
This closes #12358.
|
|
This prevents a slightly ugly backtrace instead of a genuinely
useful report.
|
|
|
|
This shouldn't be required, but adding it is harmless and indicates
our intent cleanly.
|
|
This avoids any errors of the form:
invalid byte sequence in US-ASCII
|
|
|
|
We've always had an implicit rule in macro-definitions, that they
end with a period. This meant that the first line is valid:
FOO is bar.vm.bytemark.co.uk.
However we'd expect this to fail:
FOO is bar.vm.bytemark.co.uk
A similar issue would arise if a macro-definition involved more than
one host, only the first would be valid.
We've fixed this now, such that the trailing period is optional.
|
|
|
|
This allows better alerting.
|
|
In the past we needed to write:
must run tcp on 3306.
Now we can add the "port" to match the rest of the tests:
must run tcp on port 3306.
|
|
|
|
This allows you to write the following (identical) tests:
foo must run FTP.
foo must run ftp.
foo must run FtP.
This is mostly a neatness update.
|
|
This will ensure that it uses the correct port.
|
|
This just does a TCP-connection to port 3389.
|
|
|
|
This caters to the new systems too.
|
|
We've successfully deployed on ruby 1.8, 1.9, and 2.1
|
|
|
|
If a test fails then we sleep for a small amount of time, two seconds
by default, before repeating it.
This delay is not required for tests that explicitly disable themselves.
|
|
These are not required.
|
|
|
|
|
|
|
|
|
|
|
|
Rather than showing "2, 4, etc" we use the expansion routine
to show TEST_FAILED, TEST_PASSED, etc. This reads more neatly.
|
|
Specifically that we can get the value back as a string, and
work with the ordinal too.
|
|
We no longer test for open-relays so we don't need to run the test.
|
|
Rather than having "inverted" we now name the test-method
"inverted?" which is more ideomatic.
|
|
This update consists of two changes:
* No longer return "true" or "false" instead return "TEST_FAILED", or "TEST_SUCCEEDED".
* Removed the testing of test-inversion from the class, now it lives in the base-class where it should have done all along.
|
|
|
|
This stops warnings when running with "ruby -w".
|
|
Now tests will no longe return "true" or "false", instead they
will return "TEST_FAILED" or "TEST_PASSED". There is a third
return-value of "TEST_SKIPPED" which essentially ignores a test.
|
|
This commit introduces a major change in custodian. In the past
any test had either two results:
* return false - The test failed.
* return true - The test passed.
We're now using an enum, more or less, such that a test may return a
"skipped" result which will neither raise nor clear any alert(s).
This is useful in its own right but is being introduced specifically
to allow SSL-certificate tests to avoid raising and clearing outside
working days/hours.
This closes #10328.
This fixes #10328.
|
|
We do that with the standard "_"-prefix.
|
|
We've now prefixed our transient/unused variables with "_" to stop
any future errors about variables not being used.
|
|
We'd failed to set the default value of our loaded-marker to false,
although in-practice this wasn't a problem.
|
