Use Sentry monitor service

Added

• Use Sentry inProgress monitor check-in

Externally-scheduled sync execution

Changed

• Changed the container to run /data/yii batch/full instead of the cron service. This requires a change to the deployment of this service. Rather than using a constantly-running ECS service, it should be triggered to run on a schedule by Cloudwatch or something similar.

Removed

• Removed frontend (webhook API) and Apache server
• Removed .scrutinizer.yml
• Removed s3-expand
• Removed Vagrantfile

Sentry logging

Added

• Use Sentry for error logging
• New pull request template

removed status check of other microservices

Fixed

• Removed the check of id-broker and email-service in /site/system-status because these services have their own checks. Having these checks here was causing ECS to mark this service as unhealthy and terminate the task.
• Use a more deterministic dependency list format.
• Dependency update, including new email-service-php-client and idp-id-broker-php-client that report DNS error correctly.

Update dependencies to avoid a guzzlehttp security issue

Changed (non-breaking)

• Minor upgrades to composer dependencies

Create one error message for all update-failed users

Changed (non-breaking)

• Don't cause an email to be sent out for every error. Combine errors into one.

Improvements in error handling

Fixed

• Pass the http status code through exception handling
• Add some additional error messages

Stop sending email alerts for 503 and 504 response codes

Fixed

• Stop sending email alerts for 503 and 504 response codes

Clarify an error, update dependencies

Fixed

• Improve error log by catching and adding to an ID Store exception.
• Update dependencies
• Use a specific major version of phpMyAdmin for local development work

If syncAll() throws an exception, make sure it gets logged and emailed

Merge pull request #139 from silinternational/develop

Release 4.3.1 If syncAll() throws an exception, make sure it gets logged and emailed