Are Legacy Questions Still Relevant?
A few years ago I remember Steve asking some questions I’m wondering if you find relevant still today. One question I recall simply because of the amount of response received. The question was, “What is your Server Naming Convention, and is it important what you choose?” As many companies are moving toward hosting their services in the Cloud, or balancing their systems on a more flexible internal hardware grid, less information may be known about specific servers at the application level.
So, I’d like to raise the question again today, asking if naming conventions are still relevant? Many companies may use cloud resources while still maintaining each individual server (ala Amazon). Others may have their own hardware or virtual servers they still manage to a large degree. I’m anticipating in those environments, naming conventions may be useful.
Another question I would like to ask today is regarding server monitoring software. Let me make it simply with a fictitious scenario. What monitoring tools/techniques would you recommend to a System Administrator who has one database server, one web server and one application server to manage, and a budget of say $2,000 US Dollars?
I know what I would want; a bigger budget. I love to be able to have a product maintaining statistics of CPU Utilization, Disk Wait Times, Memory Usage, Disk Usage, Page Faults and other various metrics. Having a history of those statistics can help you when utilization changes. The lighter the impact of monitoring on the server the better.
With WMI exposed you can roll your own monitors using tools built into Windows Server, and SQL Server has many techniques for monitoring resources as well. Some of them can be a little heavy resulting in too much impact on performance. I tried to run a performance monitor the redirected output to a database. I created trace values for CPU, Memory Utilization, Disk Wait and Page Faults. For that particular system, it pretty much brought it to its knees. The server was already overwhelmed and unable to handle the trace load. In other environments I have done the same monitoring technique elsewhere with great success.
With a $2,000 budget, would you build your own, use entry level products, or simply start putting together your justification for a larger budget to get a more powerful monitoring tool like those from Quest or Red Gate?
If you wish to get into this conversation, send your response to btaylor@sswug.org.
Cheers,
Ben
$$SWYNK$$
Featured Article(s)
Combining Multiple Queries In a Single One
This is often the case when you need to query more than one table in a single query, or even issue more than one subquery to a single table, combining those subqueries in a single query. There are several types of queries that come within the above description, including joins, hierarchical queries, and subqueries. In this article, you�ll look at several examples of subqueries and set operations that you can use to combine those subqueries together in queries.
Featured White Paper(s)
Achieve an astounding return on investment with Toad® for Oracle
read more)
Featured Script
alter_list_of_tables.sql
Script used to dynamically alter a list of tables(in this case add an id column and pk where there was none)… (read more)