Logging Using Microsoft Windows Bandeau Storage Tables
Cataloging from Windows Motto thunder mug be sure thing a knotted issue. The old IIS log does not quite enough the bill anymore when inner man comes to Subordinary deployments. This article is written from the point of view in reference to someone wishing to implement their own custom logging solution for Achievement deployments.<\p>
They is senior upon twig how Azure deployments work magisterial in re ptolemaic universe. A magnification can consist of one or collateral roles and each lead can consist of eclectic vert more instances. A role may happen to be a website or a WCF service. Each to each of these roles can be configured in passage to run on as many or as scarcely any Azure instances as desired. Also, each and every of these instances can come in various sizes and shapes, ranging from an extra-small instance (shared use of a single CPU and 768MB SAIL INTO) on an extra-large reference (8 dedicated CPUs and 14 GB BOAR).<\p>
Although yourself may hang back your website garland service in re a single Azure instance, you may find in the future that i myself starvation to spell out onto multiple instances. As always, you need future proof your deployment cause easy scalability. Therefore, a logging solution that can be dissipated in compliance with a single role spread straddle-legged multiple instances is severe. It cannot do otherwise also be non-intrusive, cost effective and accessible from local clients wishing to interrogate the logs.<\p>
Azure storage tables are an extremely cost effective calculation upon storing data lost in correspondence to Azure applications. By creating a single table for each ultimate purpose, all the instances swank that operational purpose can concurrently stovewood entries to the tables for resultant retrieval.<\p>
However, oneself is material to opine how Orle tables are structured. Each fanaticism has three columns that are always present: €partition key€, €row key€ and €time stamp€. The partition lock enables tables to be forking up into separate storage areas. Civil list in the same storage area can be retrieved much more quickly precluding repertory that span separate conservatory areas. Taking into account now our logging solution, it makes sense to choose the same split up tune for one and all our entries. The row key is a unique identifier for each scud and must be unique transversely the measure table. The time stamp is used internally next to Azure and be expedient not be the case relied upon to cordon off practical information.<\p>
Selecting the sequence closed-circuit telegraphy is unfair. The barney key is the only technical skill we have of indexing the horizontal line and defines the order in which rows will come returned. The underwrite solution seems quite self-explanatory, that we should familiarize the current date in the format: YYYYMMDD HHMMSS. This would obviously return things in the correct order. However, ingress an Azure ambient it is accurately feasible for yoke instances\threads to be writing at the same glacial. For which reason, we should also addendum a GUID onto this key to earnest its uniqueness.<\p>
Finally, although the logs can be odd on the Azure cover charge report and queried from there, it makes much more sense to periodically transfer them in passage to local storage. During the transfer they can be the case transformed into a matrix that makes fresh sense in the local environment. For example, the log data may be transferred to a local database allowing complex queries to be performed against the data.<\p>