@LongRunningScript, search logging, warning providers improvements

[labkey-adam]

Tasks

• dev
• code review
• manual testing?
• user ed handoff?
• teamcity review and merge

---

Rationale

Several minor improvements to make admins' (and our) lives easier:

1. New @LongRunningScript SQL upgrade script "annotation" that indicates the script may take a long time to execute. A logged message gives the reason provided in the annotation and strongly discourages admins from shutting down the server mid-upgrade. (Inspired by a recent encounter with an impatient admin who kept shutting down their server during the INT to BIGINT migration.) Update that BIGINT script with the annotation.
2. Remove the SqlScriptManager hasBeenRun() and update() methods. These are pointless since SQL scripts are never re-run or updated in the database.
3. Log the reason why the server is clearing LastIndexed for all document providers. We already log a reason for deleteIndex(), but I've seen cases where the server clears LastIndexed without it being obvious (to me) why.
4. Clear the "static" warnings whenever a new WarningProvider is registered. This allows providers that get registered after the first request the ability to still provide static warnings. Update the MailHelper warning provider to take advantage of this.
5. No need to log warnings on production deployments when an optional feature flag name doesn't conform to property name rules
6. In dev mode, detect and log orphaned attachments just before every container is deleted. This may alert us to code paths that delete objects without deleting associated attachments.