* scape strings before executing scripts in order to prevent exploits: django templates automatically scapes things. Most important is to ensuer that all escape ' to "
* Optimize SSH: pool, `UseDNS no`
* Don't store passwords and other service parameters that can be changed by the services i.e. mailman, vps etc. Find an execution mechanism that trigger `change_password()`
* abort transaction on orchestration when `state == TIMEOUT` ?
* filter and other user.is_main refactoring
* use format_html_join for orchestration email alerts
* generic form for change and display passwords and crack change password form
* enforce an emergency email contact and account to contact contacts about problems when mailserver is down
* add `BackendLog` retry action
* move invoice contact to invoices app?
* wrapper around reverse('admin:....') `link()` and `link_factory()`
Remember that, as always with QuerySets, any subsequent chained methods which imply a different database query will ignore previously cached results, and retrieve data using a fresh database query.
* profile select_related vs prefetch_related
* use HTTP OPTIONS instead of configuration endpoint, or rename to settings?
* Timezone awareness on monitoring system (reading server-side logs with different TZ than orchestra) maybe a settings value? (use UTC internally, timezone.localtime() when interacting with servers)
* EMAIL backend operations which contain stderr messages (because under certain failures status code is still 0)
* DOCUMENT: orchestration.middleware: we need to know when an operation starts and ends in order to perform bulk server updates and also to wait for related objects to be saved (base object is saved first and then related)
orders.signales: we perform changes right away because data model state can change under monitoring and other periodik task, and we should keep orders consistency under any situation.
dependency collector with max_recursion that matches the number of dots on service.match and service.metric
