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.

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

Bug fixes pr devel #401

Merged
merged 222 commits into from
Jan 18, 2024
Merged

Bug fixes pr devel #401

merged 222 commits into from
Jan 18, 2024

Conversation

gregharvey
Copy link
Member

No description provided.

EmlynK and others added 30 commits January 20, 2022 10:13
…cs with backup (#94)

* Change location of nightly backup script and delegate the cron that runs it to the deploy server, if required (for ASGs).

* Set up nightly backup crons in separate files in /etc/cron.d

* Can't put site cron files in /etc/cron.d because the deploy user doesn't have perms.

* Try and add the ability to sync a site using a nightly backup instead of a fresh DB dump.

* Use Ansible modules to look up RDS host and to copy the nightly backup into place.

* Delegate PATH setup in db backup cron to localhost.

* Shell bad. Command good. But makes it convoluted. Oh well.

* Used wrong database name in source database copy.

* Try and disable the ReplaceUnhealthy auto scale process during syncs.

* Fix Drupal cron roles when deferring to deploy server.
* Including an ami.yml for packing AMIs on build finale.

* New api_call role, focused on GitLab for now.

* Optionally trigger an infra build with an API call.
…104)

* Making the MySQL dump command for routine back-ups less aggressive.

* Making max_allowed_packets a variable we can set.

* Adding dump command to a re-usable central var.

* Renaming _init var so it's easier to distinguish in the code.
* Making the MySQL dump command for routine back-ups less aggressive.

* Making max_allowed_packets a variable we can set.

* Adding dump command to a re-usable central var.

* Renaming _init var so it's easier to distinguish in the code.

* Defaults file must be the first param for mysqldump.
* Don't recurse through site directory when setting permissions during config_generate step.

* Update drupal7 config_generate perms update task and use true/false instead of yes/no.
* Pass -l option in drush commands to specify site name, which should match the site folder name.

* Move -l option in drush commands to the first option instead of the last.
* Attempt to fix static credentials handling when deploying to an ASG.

* Include build_type in static password file location.

* Move to using new location for static credentials handling, but try to catch any legacy stuff too, for now.

* Add a couple of comments to legacy handling of static credentials.
* Add a files_sync role.

* Clear up some comments, add temp_dir variable and use rsync with command module instead of synchronize module, in files_sync role.

* Need trailing slash on src when syncing files to destination server so the contents get synced and not the parent directory itself.

* Use rsync instead of copy when syncing files to target server as copy is so slow.

* Ensure file sync tasks are run only once.
* Add a drupal_sync_tasks role to run Drupal specific tasks during syncs.

* Move sync database_apply files into tasks subdirectory.

* Woops, used bad role names in sync database update roles.

* Remove feature reverting from drupal7 database_apply sync role.
* Fixing GRANT query for MySQL > 8.0.

* Create what though?
* Adding SquashFS option to syncing.

* Make 'tarball' the default mount type so nothing existing breaks.

* Altering deploy_path so we can build in another location for SquashFS builds.

* Ensuring the build dir exists if doing a SquashFS build.

* Force symlink creation as the deploy dir may not exist yet.

* Running mount commands with sudo.

* Tweaking link command and destination for live links.

* Slight bug in link path handling for SquashFS.

* Checking for existing mount and using remount operation.

* Stop ce-deploy trying to delete from read-only SquashFS mount.

* Formatting error, these special jinja2 things are not filters - no spaces.

* Reloading services to make sure mounting doesn't fail.

* Picking more sensible loop var name.

* Moving to shell for mount check and fixing jinja2 filter names.

* Trying with the posix mount module instead of command.

* Working through user/sudo issues.

* Altering symlink handling slightly so we have the deploy directory always set.

* Simplifying 'when' checks.
* Use a different variable for config imports during a sync so they can be optional.

* Actually, no. Use a variable to determine if config should be imported during a sync.

* Fix some logic with config import during syncs and add a comment to explain when the task runs.
* Adding SquashFS option to syncing.

* Make 'tarball' the default mount type so nothing existing breaks.

* Altering deploy_path so we can build in another location for SquashFS builds.

* Ensuring the build dir exists if doing a SquashFS build.

* Force symlink creation as the deploy dir may not exist yet.

* Running mount commands with sudo.

* Tweaking link command and destination for live links.

* Slight bug in link path handling for SquashFS.

* Checking for existing mount and using remount operation.

* Stop ce-deploy trying to delete from read-only SquashFS mount.

* Formatting error, these special jinja2 things are not filters - no spaces.

* Reloading services to make sure mounting doesn't fail.

* Picking more sensible loop var name.

* Moving to shell for mount check and fixing jinja2 filter names.

* Trying with the posix mount module instead of command.

* Working through user/sudo issues.

* Altering symlink handling slightly so we have the deploy directory always set.

* Simplifying 'when' checks.

* Adding a revert behaviour for SquashFS.

* Making some wording a little less ambiguous.

* Fixing bad image path and adding clauses to stat check.

* Trying to make remounting SquashFS images a bit safer.

* Adding when clauses to SquashFS image path check.

* Removing revert code, as it cannot work.
…159)

* Add cache clears to Drupal deployments, before database updates and config import are run.

* Move cache clears out of meta roles and only run in database_apply role on non-initial builds.

* Add cache clear stuff to drupal7 database_apply role.
)

* Avoid leaving exponentially growing sqsh files in build locations!

* Use mount point instead of /tmp for stowing previous sqsh file.
* Avoid leaving exponentially growing sqsh files in build locations!

* Use mount point instead of /tmp for stowing previous sqsh file.

* Making sure image/tarball filenames are build specific in /tmp.
* Removing unnecessary lines in Drupal config generation.

* Drupal 7 lacked the install_command var.
* Ensuring dump directory exists on backup step.

* Acts on shared storage, so can run once.
* Supporting option to suppress reverting backups.

* Fixing namespacing in database_backup role.
* Handling the 'drush deploy' command more elegantly for Drupal 8+.

* Changing logic to AND for cache rebuild.
* Fixing linting so CI can work.

* More linting fixes.

* Downgrading Ubuntu due to Docker issues with ce-dev and latest.

* Making pipefail shell commands use /bin/bash.

* Adding pipefail code to cachetool installer block.

* Adding executable for drush install checker.

* Fixing D7 ctools behaviour to match features handling.

* Jinja2 template looking for settings.php using wrong path.

* Fixing config_generate for Drupal.

* Making host checking more robust.

* Removing variable declaration that will break host check.

* Allow us to use deploy_code with a completely custom build and do nothing.

* Make PHP cachetool install optional.

* Fixing bug introduced by failed_when - need to check for return code on mount points.

* Moving where opcache clears get called so they can be excluded if necessary.

* Fixing linting errors around white space.

* The unarchive module requires 'dest' to be a directory.

* Fixing Drupal example templates.

* Logically splitting tests and removing LGD because it's broken.

* Adding extra plugin LGD needs.

* Clearing stat and apcu cache for PHP by default.

* Adding some perms handling for Drupal 8+ config sync where deploy and web user are different.

* Rolling back APCu cache by default as the package must be installed.

* Making the remount squashfs code more defensive by using a block.

* Trying with 'throttle' keyword to restrict to one machine at a time.

* Sadly 'throttle' seems to work on a task level only, not a block level.

* Adding hash salt generation for Drupal 8+.

* Trying a different syntax for passing params to the password plugin.

* Forgot the constants import in plugin.

* Reinstating the SOPS plugin - we *do* need it.

* Making online/offline tasks use current live code/database.

* We need build paths to bring the site back online if it's a squashfs build.

* Ensuring _common.sh includes /etc/profile.d for Python venv path.

* Making linter happy.
* Fixing linting so CI can work.

* More linting fixes.

* Downgrading Ubuntu due to Docker issues with ce-dev and latest.

* Making pipefail shell commands use /bin/bash.

* Adding pipefail code to cachetool installer block.

* Adding executable for drush install checker.

* Fixing D7 ctools behaviour to match features handling.

* Jinja2 template looking for settings.php using wrong path.

* Fixing config_generate for Drupal.

* Making host checking more robust.

* Removing variable declaration that will break host check.

* Allow us to use deploy_code with a completely custom build and do nothing.

* Make PHP cachetool install optional.

* Fixing bug introduced by failed_when - need to check for return code on mount points.

* Moving where opcache clears get called so they can be excluded if necessary.

* Fixing linting errors around white space.

* The unarchive module requires 'dest' to be a directory.

* Fixing Drupal example templates.

* Logically splitting tests and removing LGD because it's broken.

* Adding extra plugin LGD needs.

* Clearing stat and apcu cache for PHP by default.

* Adding some perms handling for Drupal 8+ config sync where deploy and web user are different.

* Rolling back APCu cache by default as the package must be installed.

* Making the remount squashfs code more defensive by using a block.

* Trying with 'throttle' keyword to restrict to one machine at a time.

* Sadly 'throttle' seems to work on a task level only, not a block level.

* Adding hash salt generation for Drupal 8+.

* Trying a different syntax for passing params to the password plugin.

* Forgot the constants import in plugin.

* Reinstating the SOPS plugin - we *do* need it.

* Making online/offline tasks use current live code/database.

* We need build paths to bring the site back online if it's a squashfs build.

* Ensuring _common.sh includes /etc/profile.d for Python venv path.

* Making linter happy.

* Using the drush_live_bin var in sync jobs for Drupal 8+.
* Fixing linting so CI can work.

* More linting fixes.

* Downgrading Ubuntu due to Docker issues with ce-dev and latest.

* Making pipefail shell commands use /bin/bash.

* Adding pipefail code to cachetool installer block.

* Adding executable for drush install checker.

* Fixing D7 ctools behaviour to match features handling.

* Jinja2 template looking for settings.php using wrong path.

* Fixing config_generate for Drupal.

* Making host checking more robust.

* Removing variable declaration that will break host check.

* Allow us to use deploy_code with a completely custom build and do nothing.

* Make PHP cachetool install optional.

* Fixing bug introduced by failed_when - need to check for return code on mount points.

* Moving where opcache clears get called so they can be excluded if necessary.

* Fixing linting errors around white space.

* The unarchive module requires 'dest' to be a directory.

* Fixing Drupal example templates.

* Logically splitting tests and removing LGD because it's broken.

* Adding extra plugin LGD needs.

* Clearing stat and apcu cache for PHP by default.

* Adding some perms handling for Drupal 8+ config sync where deploy and web user are different.

* Rolling back APCu cache by default as the package must be installed.

* Making the remount squashfs code more defensive by using a block.

* Trying with 'throttle' keyword to restrict to one machine at a time.

* Sadly 'throttle' seems to work on a task level only, not a block level.

* Adding hash salt generation for Drupal 8+.

* Trying a different syntax for passing params to the password plugin.

* Forgot the constants import in plugin.

* Reinstating the SOPS plugin - we *do* need it.

* Making online/offline tasks use current live code/database.

* We need build paths to bring the site back online if it's a squashfs build.

* Ensuring _common.sh includes /etc/profile.d for Python venv path.

* Making linter happy.

* Using the drush_live_bin var in sync jobs for Drupal 8+.

* Fixing database dump revert.

* Fixing gunzip command.

* Having to do a double step to unpack SQL file in /tmp.
* Adding container push and build code.

* Fixing some minor issues.

* Docs updates and template flexibility improvement.

* Creating separate registry name for handling docker push.

* Trying to get the registry name right.

* Adding AWS stack components to ECS builds.

* Force a docker logout.

* Need to delegate docker commands to the controller.

* Trying deleting the credentials file instead.

* Fixing var name in subnet.yml.

* Missing some ACM variables.

* Allowing us to set wait timeout for TG creation.

* Uncommenting cpu and memory vars for ECS with Fargate.

* Missed a variable in ELB creation.

* Couple of linting fixes.

* Force building and tagging of containers as an option.

* Improving docs for deploy_container role.

* Providing container description example for ECS.

* Adding 'force' option to docker builds.

* Revert "Adding 'force' option to docker builds."

This reverts commit 5574faa.

* Adding options to force ECS component refreshes.

* Allowing users to toggle public IP and execute command mode.

* The ALB needs to be on public subnets, not the same private ones as the cluster.

* Updating documentation.

* Updating ECS docs.

* New deploy_container destroy action.

* Documentation update.

* Adding a note to docs about the container destroy action.
@tymofiisobchenko tymofiisobchenko merged commit 6ba4fa3 into devel Jan 18, 2024
3 of 5 checks passed
@tymofiisobchenko tymofiisobchenko deleted the bug_fixes-PR-devel branch January 18, 2024 08:19
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

5 participants