Christos Polydorou

A good algorithm is not always the solution to a given problem. In most cases, a mechanism to save the data used by the algorithm is required, in order not only to organise the data but also optimise its execution. Such mechanisms are called Data Structures and you can think of them as small databases. Data Structures not only hold data, but also expose methods to access and update it. My DataStructures module contains the most frequently used data structures, implemented for Powershell objects in C#. At the time this post is written, the module contains the stack and queue structures, but more are on the way. Stack The idea behind the stack data structure is to save objects in order to be processed later in the execution of an algorithm. The below figure shows the way stacks work: As we insert items in the stack, the items already in it are pushed to the bottom. When removing an item, the item to be removed is the most recent one, hence the alternate naming LIFO - Last In First Out...

The design of the latest versions of On Premises Exchange server has been more and more close to the Cloud version. One of the major changes was the introduction of the Database Availability Groups, that are similar to the SQL server availability groups in concept and allow us to have more than one copies of a mailbox database. This increases the availability of the mailbox service but introduces cost for the required storage. Although storage is cheap these days, there's no need to invest much in it for Exchange since the system will recover from a failed disk (always talking about the volumes holding the database files). However, you should always plan the hardware and system requirements using the Exchange Requirements Calculator. A hardware, software or network event may lead to missed logs between the database copies and result in copy and/or replay queue length. Depending on the time taken to resolve the issue and restore the connectivity, you may have to resume the m...