The minitest handler doesn't know you don't have tests written for
those cookbooks. It's looping through all of the cookbooks for the
recipes in the node's current run list and attempting to pull down
test and support directories from all of them. These resources should
be set to "ignore on failure," so it shouldn't cause your Chef run to
bomb out if they're not there.
At the end of the run, the handler will load all of the tests that did make it to the node and run them. If any of those tests fail,
it will flag your run as failed and your exception handlers will then
run.
If you find the behavior of the handler unsatisfying and find a better
way to pull down tests from all cookbooks, I'm sure the maintainers of
the minitest handler would love to see your improvement in the form of
a pull request!
On Mon, Oct 8, 2012 at 11:51 PM, Sachin Sagar Rai millisami@gmail.com wrote:
On Tuesday, October 9, 2012 at 11:41 PM, steve . wrote:
The minitest handler doesn't know you don't have tests written for
those cookbooks. It's looping through all of the cookbooks for the
recipes in the node's current run list and attempting to pull down
test and support directories from all of them. These resources should
be set to "ignore on failure," so it shouldn't cause your Chef run to
bomb out if they're not there.
At the end of the run, the handler will load all of the tests that did make it to the node and run them. If any of those tests fail,
it will flag your run as failed and your exception handlers will then
run.
If you find the behavior of the handler unsatisfying and find a better
way to pull down tests from all cookbooks, I'm sure the maintainers of
the minitest handler would love to see your improvement in the form of
a pull request!
I wrote some of that code, it just checks whether any of the recipes
in the run_list also have tests. If it were smarter it would allow you
to filter for only those tests that you want run. There currently is
no way to do that though I would love to add that functionality
On Tuesday, October 9, 2012 at 11:41 PM, steve . wrote:
The minitest handler doesn't know you don't have tests written for
those cookbooks. It's looping through all of the cookbooks for the
recipes in the node's current run list and attempting to pull down
test and support directories from all of them. These resources should
be set to "ignore on failure," so it shouldn't cause your Chef run to
bomb out if they're not there.
At the end of the run, the handler will load all of the tests that did make it to the node and run them. If any of those tests fail,
it will flag your run as failed and your exception handlers will then
run.
If you find the behavior of the handler unsatisfying and find a better
way to pull down tests from all cookbooks, I'm sure the maintainers of
the minitest handler would love to see your improvement in the form of
a pull request!
On Mon, Oct 8, 2012 at 11:51 PM, Sachin Sagar Rai millisami@gmail.com
wrote:
I guess you could put a whitelist/blacklist set of tests (or maybe
regexes?) into the node data, namespaced appropriately for the
minitest handler... the handler has access to this data during its
run, so it would be able to pick that up and filter its cookbook file
walk appropriately.
Anyone else want that functionality? I mean, I could see at least
some use for it, though for my current work setup it's not
critical-path...
I wrote some of that code, it just checks whether any of the recipes
in the run_list also have tests. If it were smarter it would allow you
to filter for only those tests that you want run. There currently is
no way to do that though I would love to add that functionality
On Tuesday, October 9, 2012 at 11:41 PM, steve . wrote:
The minitest handler doesn't know you don't have tests written for
those cookbooks. It's looping through all of the cookbooks for the
recipes in the node's current run list and attempting to pull down
test and support directories from all of them. These resources should
be set to "ignore on failure," so it shouldn't cause your Chef run to
bomb out if they're not there.
At the end of the run, the handler will load all of the tests that did make it to the node and run them. If any of those tests fail,
it will flag your run as failed and your exception handlers will then
run.
If you find the behavior of the handler unsatisfying and find a better
way to pull down tests from all cookbooks, I'm sure the maintainers of
the minitest handler would love to see your improvement in the form of
a pull request!
On Mon, Oct 8, 2012 at 11:51 PM, Sachin Sagar Rai millisami@gmail.com
wrote:
i think cucumber and rspec use tags "@foobar" to filter the tests to
run, i would try to support that filtering mechanism rather than
whitelist/blacklist
I guess you could put a whitelist/blacklist set of tests (or maybe
regexes?) into the node data, namespaced appropriately for the
minitest handler... the handler has access to this data during its
run, so it would be able to pick that up and filter its cookbook file
walk appropriately.
Anyone else want that functionality? I mean, I could see at least
some use for it, though for my current work setup it's not
critical-path...
I wrote some of that code, it just checks whether any of the recipes
in the run_list also have tests. If it were smarter it would allow you
to filter for only those tests that you want run. There currently is
no way to do that though I would love to add that functionality
On Tuesday, October 9, 2012 at 11:41 PM, steve . wrote:
The minitest handler doesn't know you don't have tests written for
those cookbooks. It's looping through all of the cookbooks for the
recipes in the node's current run list and attempting to pull down
test and support directories from all of them. These resources should
be set to "ignore on failure," so it shouldn't cause your Chef run to
bomb out if they're not there.
At the end of the run, the handler will load all of the tests that did make it to the node and run them. If any of those tests fail,
it will flag your run as failed and your exception handlers will then
run.
If you find the behavior of the handler unsatisfying and find a better
way to pull down tests from all cookbooks, I'm sure the maintainers of
the minitest handler would love to see your improvement in the form of
a pull request!
On Mon, Oct 8, 2012 at 11:51 PM, Sachin Sagar Rai millisami@gmail.com
wrote:
But still wondering where to put that 'ignore on failure' line? In tests?
Where to put it exactly?
What it comes down to, is you don't have to do anything. It's already
ignoring the cookbooks with no tests. The output you see just happens to
be overly verbose and ugly, but it's not affecting your run at all.