New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Test suite settings should take priority over settings passed by POST #684
Comments
I am against this. I think POST requests should have priority. Not everyone has write access to test suites so if someone is developing a test and want to experiment with different settings she must use API call. Now you have a point, that when there is a dependent test it doesn't make sense to override its variables. So one solution would be to add this logic to iso controller. |
The number of disks os-autoinst prepares depends on NUMDISKS variable. So theoretically you can use HDD_2_URL in POST and set HDD_1=%HDD_2% in test suite settings for the parent job and keep the child job as is. |
That's true. Other solution would be to create some variable like "NOT_OVERWRITE=HDD_1". I will try your solution on Monday. |
+1 for leaving the priority as it is. |
+1 for leaving the prio as it is |
So using |
It is hacky, true. |
As suggested by @aaannz in os-autoinst#684, allow specifying a variable as `+VARIABLE` to signify that it should not be overridden by another setting of `VARIABLE` later in the usual precedence order. For instance, usually if the same variable is set in both a product and a test suite, the test suite setting wins; with this change, if the product specified `+VARIABLE` and the test suite `VARIABLE`, the product setting would win. One major use for this could be to prevent an `HDD_1` set from an ISO post request overriding the value set in templates for a test which uses a disk image uploaded by another test. If multiple things specify `+VARIABLE`, I believe the one that comes highest in the normal precedence order should win out. We could, of course, extend this to give `++VARIABLE` even greater precedence, and so on...:)
As suggested by @aaannz in os-autoinst#684, allow specifying a variable as `+VARIABLE` to signify that it should not be overridden by another setting of `VARIABLE` later in the usual precedence order. For instance, usually if the same variable is set in both a product and a test suite, the test suite setting wins; with this change, if the product specified `+VARIABLE` and the test suite `VARIABLE`, the product setting would win. One major use for this could be to prevent an `HDD_1` set from an ISO post request overriding the value set in templates for a test which uses a disk image uploaded by another test. If multiple things specify `+VARIABLE`, I believe the one that comes highest in the normal precedence order should win out. We could, of course, extend this to give `++VARIABLE` even greater precedence, and so on...:)
As suggested by @aaannz in os-autoinst#684, allow specifying a variable as `+VARIABLE` to signify that it should not be overridden by another setting of `VARIABLE` later in the usual precedence order. For instance, usually if the same variable is set in both a product and a test suite, the test suite setting wins; with this change, if the product specified `+VARIABLE` and the test suite `VARIABLE`, the product setting would win. One major use for this could be to prevent an `HDD_1` set from an ISO post request overriding the value set in templates for a test which uses a disk image uploaded by another test. If multiple things specify `+VARIABLE`, I believe the one that comes highest in the normal precedence order should win out. We could, of course, extend this to give `++VARIABLE` even greater precedence, and so on...:)
As suggested by @aaannz in os-autoinst#684, allow specifying a variable as `+VARIABLE` to signify that it should not be overridden by another setting of `VARIABLE` later in the usual precedence order. For instance, usually if the same variable is set in both a product and a test suite, the test suite setting wins; with this change, if the product specified `+VARIABLE` and the test suite `VARIABLE`, the product setting would win. One major use for this could be to prevent an `HDD_1` set from an ISO post request overriding the value set in templates for a test which uses a disk image uploaded by another test. If multiple things specify `+VARIABLE`, I believe the one that comes highest in the normal precedence order should win out. We could, of course, extend this to give `++VARIABLE` even greater precedence, and so on...:)
As suggested by @aaannz in os-autoinst#684, allow specifying a variable as `+VARIABLE` to signify that it should not be overridden by another setting of `VARIABLE` later in the usual precedence order. For instance, usually if the same variable is set in both a product and a test suite, the test suite setting wins; with this change, if the product specified `+VARIABLE` and the test suite `VARIABLE`, the product setting would win. One major use for this could be to prevent an `HDD_1` set from an ISO post request overriding the value set in templates for a test which uses a disk image uploaded by another test. If multiple things specify `+VARIABLE`, I believe the one that comes highest in the normal precedence order should win out. We could, of course, extend this to give `++VARIABLE` even greater precedence, and so on...:)
As suggested by @aaannz in #684, allow specifying a variable as `+VARIABLE` to signify that it should not be overridden by another setting of `VARIABLE` later in the usual precedence order. For instance, usually if the same variable is set in both a product and a test suite, the test suite setting wins; with this change, if the product specified `+VARIABLE` and the test suite `VARIABLE`, the product setting would win. One major use for this could be to prevent an `HDD_1` set from an ISO post request overriding the value set in templates for a test which uses a disk image uploaded by another test. If multiple things specify `+VARIABLE`, I believe the one that comes highest in the normal precedence order should win out. We could, of course, extend this to give `++VARIABLE` even greater precedence, and so on...:)
commit a3709d0 Author: Adam Williamson <adamw@happyassassin.net> AuthorDate: Sun Jan 29 13:29:09 2017 +0100 Commit: Stephan Kulow <coolo@kde.org> CommitDate: Sun Jan 29 13:29:09 2017 +0100 Allow override of the usual setting precedence order (#1200) As suggested by @aaannz in #684, allow specifying a variable as `+VARIABLE` to signify that it should not be overridden by another setting of `VARIABLE` later in the usual precedence order. For instance, usually if the same variable is set in both a product and a test suite, the test suite setting wins; with this change, if the product specified `+VARIABLE` and the test suite `VARIABLE`, the product setting would win. One major use for this could be to prevent an `HDD_1` set from an ISO post request overriding the value set in templates for a test which uses a disk image uploaded by another test. If multiple things specify `+VARIABLE`, I believe the one that comes highest in the normal precedence order should win out. We could, of course, extend this to give `++VARIABLE` even greater precedence, and so on...:)
Fedora for ARM is distributed not by installation ISO, but by already preinstalled disk image. When I want to schedule our ARM tests, I'm doing so while setting
HDD_1_URL
, but notISO
. I have one test that boots from this disk, creates user/sets password with our "initial-setup" utility and then saves disk by settingSTORE_HDD_1
. Then I have another ARM test where I want to use that saved disk, but I cannot, becauseHDD_1
set on test gets overwritten by parameter passed during test scheduling.I tried to overcome this by setting
HDD_ARM=http://url/to/hdd
during POST,HDD_1_URL=%HDD_ARM%
on first test and correctly settingHDD_1
on second test, but_URL
gets resolved only for parameters set by POST (soHDD_1
wasn't created and disk wasn't downloaded). I also tried to setASSET_1_URL=http://url/to/hdd
during POST,HDD_1=%ASSET_1%
on first test and correctly settingHDD_1
on second test and that almost worked (ASSET_1
was created and disk was downloaded), but then disk was placed inother/
directory, not inhdd/
directory (so it couldn't find it).I cannot come up with mean how to solve this problem (resolving
_URL
for all parameters would be enough), but I think that parameters set on specific test should take precedence before "generic" parameters passed during POST for all tests.Rationale is that when I'm passing parameters that are the same for all tests planned by one POST, I should be able to override parameters for specific tests.
The text was updated successfully, but these errors were encountered: