Architecture Tactics - Revision history

Suhridk at 17:14, 23 April 2012

2012-04-23T17:14:36Z

2012-04-22T16:20:09Z

State resynchronization

2012-04-22T15:45:41Z

Exceptions

2012-04-22T15:41:54Z

Availability Tactics

2012-04-01T12:33:17Z

Design time Tactics

2012-04-01T12:24:17Z

Recovering from attacks

2012-04-01T11:54:45Z

= Preparation and Repair Tactics

2012-04-01T11:54:00Z

Detecting attacks

2012-04-01T11:45:58Z

Resisting attacks

2012-04-01T11:37:57Z

Static Scheduling

← Older revision		Revision as of 17:14, 23 April 2012
Line 332:		Line 332:

	* Built in monitors: Record events when monitoring states have been activated. Increased visibility into activities of the component.		* Built in monitors: Record events when monitoring states have been activated. Increased visibility into activities of the component.
		+
		+	[[Category:SYAR]]

← Older revision		Revision as of 16:20, 22 April 2012
Line 83:		Line 83:

	* Restored component must have its state upgraded before return to service.		* Restored component must have its state upgraded before return to service.
		+	* Passive and active redundancy tactics require this.
	* Ideal approach to update the state is a single atomic message. Incremental state upgrades lead to complicated software.		* Ideal approach to update the state is a single atomic message. Incremental state upgrades lead to complicated software.

@@ Line 37: / Line 37: @@
 === Exceptions ===
-* Exceptions encountered during an exception.
+* Exceptions encountered during operation.
 * Exception handler is invoked which typically executes in the same process that introduced the exception.

← Older revision		Revision as of 15:41, 22 April 2012
Line 15:		Line 15:

	* All approaches to maintaining availability involve some type of redundancy, some type of health monitoring and some type of recovery when a failure is detected.		* All approaches to maintaining availability involve some type of redundancy, some type of health monitoring and some type of recovery when a failure is detected.
		+	* Faults cause failures. Availability tactics focus on dealing with faults.
	* Availability tactics involve- Fault detection, fault recovery and fault prevention.		* Availability tactics involve- Fault detection, fault recovery and fault prevention.

← Older revision		Revision as of 12:33, 1 April 2012
Line 318:		Line 318:
	* Separate UI from rest of application: Allows UI developers to frequently change the UI and maintain code separately.		* Separate UI from rest of application: Allows UI developers to frequently change the UI and maintain code separately.
	* Classic design pattern to implement this tactic: Model-View-Controller (MVC).		* Classic design pattern to implement this tactic: Model-View-Controller (MVC).
		+
		+	= Testability Tactics =
		+
		+	== Manage input/output ==
		+
		+	* Record/playback:Capture information crossing an interface and use it as input to the test harness.
		+	* Separate interface from implementation: Stubbing implementation allows rest of system to be tested in absence of stubbed component.
		+	* Specialize access routes/interfaces: Have specialized testing interfaces.
		+
		+	== Internal Monitoring ==
		+
		+	* Built in monitors: Record events when monitoring states have been activated. Increased visibility into activities of the component.

← Older revision		Revision as of 12:24, 1 April 2012
Line 299:		Line 299:
	** Especially maintain redundant copoes of sys admin data such as passwords, ACL's, and user profile data.		** Especially maintain redundant copoes of sys admin data such as passwords, ACL's, and user profile data.

−	* Tactics for identifying attackers ~~is to maintain~~ an audit trail. Audit can be used trace actions of attacker, support non repudiation and support system recovery.	+	* Tactics for identifying attackers: Maintain an audit trail. Audit can be used trace actions of attacker, support non repudiation and support system recovery.
		+
		+	= Usability Tactics =
		+
		+	== Runtime Tactics ==
		+
		+	* Support user initiative:
		+	** Undo, redo, aggregate: All require architectural consideration.
		+
		+	* Support system initiative:
		+
		+	* Maintain a task model: Task analysis. e.g. auto correct beginning of sentences to capital.
		+	* Maintain a model of user: e.g. Personas. How much the user's knows about the system, users behaviour in terms of expected response times. e.g. Page scrolling rate.
		+	* Maintain a model of the system: Determines expected system behaviour so that appropriate feedback can be given to user. e.g. time needed to complete certain activity.
		+
		+	== Design time Tactics ==
		+
		+	* Separate UI from rest of application: Allows UI developers to frequently change the UI and maintain code separately.
		+	* Classic design pattern to implement this tactic: Model-View-Controller (MVC).

@@ Line 43: / Line 43: @@
 * Fault recovery consists of preparing for recovery and making the actual system repair as well reintroduction of components after repair.
-=== Preparation and Repair Tactics ==
+== Preparation and Repair Tactics ==
 ==== Voting ====

← Older revision		Revision as of 11:54, 1 April 2012
Line 288:		Line 288:
	== Detecting attacks ==		== Detecting attacks ==

−	*	+	* Intrusion detection systems. e.g. compare network traffic patterns against known ones.
		+	* Artificial immune systems
		+	* Set traps
		+
		+	== Recovering from attacks ==
		+
		+	* Can be divided into restoring state and identifying attacker.
		+
		+	* Recovering state: related to availability tactics.
		+	** Especially maintain redundant copoes of sys admin data such as passwords, ACL's, and user profile data.
		+
		+	* Tactics for identifying attackers is to maintain an audit trail. Audit can be used trace actions of attacker, support non repudiation and support system recovery.

← Older revision		Revision as of 11:45, 1 April 2012
Line 284:		Line 284:
	# Maintain integrity: Redundant information encoded in it - e.g. checksums, hash results.		# Maintain integrity: Redundant information encoded in it - e.g. checksums, hash results.
	# Limit exposure: Attacks typically all data & services on a single host. Architect can design allocation of services to hosts so that limited services are available on each host.		# Limit exposure: Attacks typically all data & services on a single host. Architect can design allocation of services to hosts so that limited services are available on each host.
		+	# Limit access: Limit access from sources using firewalls. Establish DMZ. DMZ is a subnet that exists between the firewall protecting the internal LAN and the wider internet. Hosts within DMZ have limited connectivity to internal hosts while communicating with external hosts.
		+
		+	== Detecting attacks ==
		+
		+	*

← Older revision		Revision as of 11:37, 1 April 2012
Line 270:		Line 270:

	* Sequence of assignment of resources is determined offline.		* Sequence of assignment of resources is determined offline.
		+
		+	= Security Tactics =
		+
		+	* Can be divided into three types of tactics : resting attacks (e.g. lock), detecting attacks (e.g. sensor), recovering from attacks (e.g. insurance).
		+
		+	== Resisting attacks ==
		+
		+	* Address the requirements of security of a system.
		+
		+	# Authenticate users: Users are who they claim to be. Passwords etc.
		+	# Authorize users: Authenticated user has the rights to access and modify either data or services. Access Control Systems etc.
		+	# Maintain data confidentiality: Encryption, public key authentication.
		+	# Maintain integrity: Redundant information encoded in it - e.g. checksums, hash results.
		+	# Limit exposure: Attacks typically all data & services on a single host. Architect can design allocation of services to hosts so that limited services are available on each host.