Skip to content
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.

Sign up for GitHub

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

Jump to bottom

Running from archive still downloads scripts from the internet #887

Closed
mstoykov opened this issue Jan 3, 2019 · 0 comments · Fixed by #1059
Closed

Running from archive still downloads scripts from the internet #887

mstoykov opened this issue Jan 3, 2019 · 0 comments · Fixed by #1059
Labels
bug

Comments

@mstoykov
Copy link
Contributor

mstoykov commented Jan 3, 2019

When you have something like

import faker from 'cdnjs.com/libraries/Faker'

k6 will load it from https://cdnjs.com/libraries/Faker and than put it in the archive when you run k6 archive. The problem is that then when you k6 run archive.tar it will still go get it from the url.
@na-- na-- added the bug label Jan 3, 2019
This was referenced Jun 7, 2019
Closed
Closed
mstoykov added a commit that referenced this issue Jun 26, 2019
@mstoykov
Rewrite script/files loading to be be url based
aa4b090 
This includes also change to the archive file structure. Now instead of
separating files by whether they are scripts or no, they are separated
based on whether their URI scheme.

Through the (majority) of k6 now instead of simple strings a url.URL is
used to idenitify files(scripts or otherwise).

This also means that all imports or `open` can have schemes.
Previously remote modules were supported specifically without scheme.
With this change we specifically prefer if they are with a scheme. The
old variant is supported but logs a warning.
Additionally if remote module requires a relative/absolute path that doesn't have a
scheme it is relative/absolute given the remote module url.

Because of some of the changes, now caching is done through a afero.Fs
instead of additional map. This also fixes not laoding remotely imported
files from an archive, but instead requesting them again.

fixes #1037, closes #838, fixes #887 and fixes #1051
@mstoykov mstoykov mentioned this issue Jun 26, 2019
Merged
mstoykov added a commit that referenced this issue Jul 31, 2019
@mstoykov
Rewrite script/files loading to be be url based (#1059)
ea8384f 
This also includes a change to the archive file structure. Now instead of
separating files by whether they are scripts or not, they are separated
based on their URI scheme.

Throughout the (majority) of k6 now, instead of simple strings, a url.URL is
used to identify files (scripts or otherwise).
This also means that all imports can have schemes. Previously remote modules
were supported specifically without a scheme. With this change k6 prefers if
they are with a scheme. The old variant is supported, but logs a warning.
Additionally if remote module requires a relative/absolute path that doesn't
have a scheme, it is relative/absolute given the remote module url.

Because of some of the changes, now caching is done through a afero.Fs
instead of additional maps. This also fixes the remotely imported files not 
properly loading from an archive, but instead requesting them again. This is
supported with old archives and for github/cdnjs's shortcut loaders. 

For future-proof reasons the archive now also records the GOOS value of the 
k6 that generated it. 

Anonymize case insensitively as windows paths are (usually) case insensitive. 

fixes #1037, closes #838, fixes #887 and fixes #1051
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
bug
Projects
None yet
Milestone
No milestone
Development

Successfully merging a pull request may close this issue.

Rewrite script/files loading to be be url based grafana/k6
2 participants
@mstoykov @na--