| Age | Commit message (Collapse) | Author |
|---|
|
|
|
|
|
|
|
|
|
|
|
This commit lets us use the bleeding-edge gn-uploader from the
gn-uploader channel.
|
|
The uploader container is for testing that the gn-uploader package
builds and the application is working correctly. It sets up a service
that allows users to upload data.
It does need the other GeneNetwork services, but does not need the
latest and greatest. This commit uses the default genenetwork2
package (which will mostly be stable): thus there are fewer (if any)
issues getting the other services to run.
|
|
|
|
|
|
|
|
source code for development. See
topics/systems/debug-and-developing-code-with-genenetwork-system-container.gmi
|
|
|
|
* genenetwork/services/genenetwork.scm (<gn-uploader-configuration>):
New type.
(gn-uploader-activation, gn-uploader-gunicorn-app,
gn-uploader-nginx-server-block): New functions.
(gn-uploader-service-type): New variable.
* uploader.scm, uploader-deploy.sh: New files.
Signed-off-by: Arun Isaac <arunisaac@systemreboot.net>
|
|
|
|
|
|
|
|
Not everyone has GUIX_PROFILE set in their environment. Inferring the
profile from the guix executable is more reliable.
* genenetwork-development.scm (channels-scm-gexp): Infer profile from
guix executable used to build container.
|
|
|
|
gn-auth runs as the gunicorn-gn-auth user, not the genenetwork user.
* genenetwork/services/genenetwork.scm (genenetwork-activation): Let
the gunicorn-gn-auth user own the auth database.
Reported-by: Frederick M. Muriithi <fredmanglis@protonmail.com>
|
|
This is useful to enable building URIs with the correct scheme in the
application. These URIs are important when redirection to and from
external services such as the gn-auth service.
|
|
|
|
*
production.scm (operating-system)[services]{genenetwork-service-type}:
Remove auth-db field.
* production-deploy.sh (container_script): Mount /var/genenetwork.
|
|
|
|
* fallback.scm: New file.
|
|
* genenetwork/services/genenetwork.scm (genenetwork-gunicorn-apps):
Set AUTH_SERVER_URL on production genenetwork2.
|
|
*
genenetwork/services/genenetwork.scm (<genenetwork-configuration>)[gn-auth-server-name]:
New field.
*
genenetwork/services/genenetwork.scm (genenetwork-nginx-server-block):
Rename to ...
(genenetwork-nginx-server-blocks): ... this. Return list of reverse
proxy blocks including one for gn-auth.
* genenetwork/services/genenetwork.scm (genenetwork-service-type): Use
genenetwork-nginx-server-blocks instead of
genenetwork-nginx-server-block.
*
production.scm (operating-system)[services]{genenetwork-service-type}:
Add gn-auth-server-name.
|
|
* genenetwork/services/genenetwork.scm (genenetwork-activation): Set
owner-only permissions on secrets files.
|
|
* genenetwork/services/genenetwork.scm (genenetwork-activation): Set
ownership of secrets files.
* production-deploy.sh: Mount secrets files read-write.
|
|
* genenetwork/services/genenetwork.scm: Import gn-auth from (gn
packages genenetwork).
*
genenetwork/services/genenetwork.scm (<genenetwork-configuration>)[gn-auth,
gn-auth-port, gn-auth-secrets]: New fields.
* genenetwork/services/genenetwork.scm (genenetwork-gunicorn-apps):
Add gn-auth gunicorn app.
|
|
* genenetwork-development.scm: Bump copyright year for Fred.
|
|
*
genenetwork-development.scm (<genenetwork-configuration>)[gn-auth-secrets]:
New field.
* genenetwork-development.scm (gn-auth-cd-gexp): Set GN_AUTH_SECRTS
from gn-auth-secrets of <genenetwork-configuration> object.
(operating-system)[services]{genenetwork-service-type}: Set
gn-auth-secrets.
|
|
|
|
* genenetwork-development.scm (gn-auth-cd-gexp): Run with 8 workers.
|
|
* genenetwork-development.scm (genenetwork3-cd-gexp): Run with 8
workers.
|
|
With the reorganisation of the modules, and a change in the startup of
genenetwork2, the way mechanical-rob is launched has also
changed. This commit updates the launch command.
* genenetwork-development.scm (genenetwork-projects): Update module
path in launch of mechanical-rob CI job.
Signed-off-by: Arun Isaac <arunisaac@systemreboot.net>
|
|
* genenetwork-development.scm (guile-lmdb-project): New variable.
(operating-system)[services]{forge-service-type}: Add
guile-lmdb-project.
|
|
*
genenetwork-development.scm (transform-genenetwork-database-project):
Fix channel name.
|
|
guix requires "=" between argument names and values.
* production-deploy.sh: Add "=" before --load-path argument value.
|
|
* genenetwork/services/genenetwork.scm (genenetwork-gunicorn-apps):
Increase gunicorn timeout for genenetwork3 to 1200 seconds.
|
|
* genenetwork-development.scm (genenetwork3-cd-gexp): Increase
gunicorn timeout to 1200 seconds.
* genenetwork/services/genenetwork.scm (genenetwork-gunicorn-apps):
Add TODO note to increase gunicorn timeout to 1200 seconds.
|
|
* genenetwork-development.scm (genenetwork-activation): Set ownership
and permissions on secret files.
* genenetwork-development-deploy.sh: Share /etc/genenetwork/conf
instead of merely exposing it.
|
|
*
genenetwork-development.scm (<genenetwork-configuration>)[gn2-secrets,
gn3-secrets]: New fields.
* genenetwork-development.scm (genenetwork2-cd-gexp): Use gn2-secrets
configuration field.
(genenetwork3-cd-gexp): Use gn3-secrets configuration field.
(operating-system)[services]: Set gn2-secrets and gn3-secrets fields
in genenetwork configuration.
|
|
* production.scm: Import (genenetwork services genenetwork), (forge
acme), (forge nginx) and (forge socket). Import shepherd from (gnu
packages admin).
(operating-system)[sudoers-file]: Allow acme user to restart nginx.
(services): Add forge-nginx, acme and genenetwork services.
|
|
* genenetwork/services/genenetwork.scm: New file.
|
|
The default settings file in the genenetwork2 code base has been moved
from etc/default_settings.py to gn2/default_settings.py. Adapt.
* genenetwork-development.scm (genenetwork-projects,
genenetwork2-cd-gexp): Change path to default genenetwork2 settings
file.
|
|
Somehow this code, that was deleted in
aaf6750858e061e5f5740fb1edbc49b68b3c4717, was added back in
93ff9ed6fd9f2ca15cf6a8930eae26c9f5c65450.
* genenetwork-development.scm (qc-tests, qc-project): Delete
variables.
|
|
We split guix-bioinformatics into two jobs—one that only does a `guix
pull' and another that does a `guix pull' and builds all
packages. This allows us to catch failures in `guix pull'
separately. The job that builds all packages is practically always
bound to fail.
* genenetwork-development.scm (guix-bioinformatics-project): Split
guix-bioinformatics job into guix-bioinformatics and
guix-bioinformatics-all-packages jobs.
|
|
* production.scm (operating-system)[services]{mysql-service-type}: Set
auto-upgrade? to #f.
|
|
|
|
*
genenetwork-development.scm (transform-genenetwork-database-project)[repository]:
Set to local gn-transform-databases git repository.
[ci-jobs-trigger]: Do not use webhook. Reset to default.
|
