Time and ordering in distributed system
这个领域的开拓者是: Leslie Lamport:
"Time, Clocks, and the Ordering of Events in a Distributed System",[6] which received the PODC Influential Paper Award in 2000,[12]
Make it computational-ordering
一、关于order和computation之间的关系，参见工程discrete的Relation-structure-computation\Make-it-computational章节。
二、在分布式系统中，尤其需要处理time、ordering的问题
比较好的文章
在下面文章，对Distributed Computing中，time、ordering问题进行了比较好的探讨:
1、cockroachlabs Living Without Atomic Clocks
其中讨论了"The Importance of Time in Distributed Systems"。
2、uci Distributed Computing Concepts - Global Time and State in Distributed Systems
Ordering and synchronization
本节讨论ordering和synchronization之间的关系:
1) synchronization的目的是保持ordering
上述"Distributed system的disordering本质"章节的内容论证了这个观点。
2) ordering能够用于实现synchronization
上述"Ordering in zookeeper"章节的内容论证了这个观点。
Ordering and consistency
需要讨论ordering 和 consistency之间的关系，在下面文章中对此进行了描述:
cornell Distributed Systems: Ordering and Consistency
Distributed system的disordering本质
一、Distributed system是典型multiple entity model，即它是由多个不同的、独立的entity构成，这些entity的execution是独立的、随机的，这就是"distributed system的disordering本质"；但是从系统的整体而言，如果要实现computation、确定性，就需要设法消除这种disordering，显然就需要加入一定的control。control机制的目的是消除disordering，实现ordering。
二、通过 wikipedia Distributed computing 、Distributed-system-challenge 中的内容，我们已经知道，分布式系统的一个非常大的调整是:
因此，在分布式系统中就无法依赖于physical clock来进行ordering；
如何克服这些问题
使用logical clock
参见 Logical-clock 章节。
实现global clock
Google spanner: "GPS和原子钟"，cockroachlabs Living Without Atomic Clocks 中，总结如下:
One of the most surprising and inspired facets of Spanner is its use of atomic clocks and GPS clocks to give participating nodes really accurate wall time synchronization. The designers of Spanner call this “TrueTime”, and it provides a tight bound on clock offset between any two nodes in the system. This lets them do pretty nifty things!
参见Google-spanner章节。
Example
Ordering in TensorFlow
在阅读TensorFlow whitepaper的时候，其中有这样的描述:
Our implementation also sometimes inserts control dependencies to enforce orderings between otherwise independent operations as a way of, for example, controlling the peak memory usage.
Ordering in zookeeper
在 ZooKeeper: A Distributed Coordination Service for Distributed Applications 中很多关于ordering的描述:
The ZooKeeper implementation puts a premium(保证) on high performance, highly available, strictly ordered access.
The strict ordering means that sophisticated synchronization primitives can be implemented at the client.
ZooKeeper is ordered. ZooKeeper stamps(贴上时间戳) each update with a number that reflects the order of all ZooKeeper transactions. Subsequent operations can use the order to implement higher-level abstractions, such as synchronization primitives.
从上面的描述可以看出，"ordering能够用于实现synchronization"。
Sequential Consistency - Updates from a client will be applied in the order that they were sent.