Archive for January, 2011

Posted January 31st, 2011
Eyal Markovich

Application Performance Tuning and High Performance Storage

By Eyal Markovich
Bookmark/FavoritesFacebookDeliciousDiggGoogle+LinkedInTechnorati FavoritesStumbleUponTumblrTwitterShare

One of the advantages of enterprise super fast storage, such as the Kaminario K2, is that it can reduce or eliminate application tuning. Rather than tuning an application that is not performing well, you can achieve greater performance by moving the data to much faster storage. Read the rest of this entry »

Bookmark/FavoritesFacebookDeliciousDiggGoogle+LinkedInTechnorati FavoritesStumbleUponTumblrTwitterShare
Posted January 25th, 2011
Eyal Markovich

Startling I/O Performance to Drive Application and Business Value

By Eyal Markovich
Bookmark/FavoritesFacebookDeliciousDiggGoogle+LinkedInTechnorati FavoritesStumbleUponTumblrTwitterShare

In his white paper, “Startling I/O Performance to Drive Application and Business Value,” industry analyst Mark Peters from Enterprise Strategy Group (ESG), discusses the need for consistent speed that is well targeted at applications that require higher performance. Read the rest of this entry »

Bookmark/FavoritesFacebookDeliciousDiggGoogle+LinkedInTechnorati FavoritesStumbleUponTumblrTwitterShare
Posted January 19th, 2011
Eyal Markovich

Application Performance Requires Better Storage Performance

By Eyal Markovich
Bookmark/FavoritesFacebookDeliciousDiggGoogle+LinkedInTechnorati FavoritesStumbleUponTumblrTwitterShare

It is increasingly true that organizations in nearly all markets live or die by what their applications do for them. This means that their information systems must always perform well. Application performance goes beyond the fact that a system or groups of systems are working in some data center. It includes the effectiveness of the computing resources, networking resources, storage resources and, of course, administrative processes and procedures. Applications that don’t meet performance expectations can create problems for organizations. The pain might be felt in many other areas that seem to be unrelated to the organization’s information systems. Read the rest of this entry »

Bookmark/FavoritesFacebookDeliciousDiggGoogle+LinkedInTechnorati FavoritesStumbleUponTumblrTwitterShare
Posted January 12th, 2011
Eyal Markovich

Database Performance Assessments for Oracle and Microsoft SQL Server

By Eyal Markovich
Bookmark/FavoritesFacebookDeliciousDiggGoogle+LinkedInTechnorati FavoritesStumbleUponTumblrTwitterShare

The majority of applications have a backend database which enables critical data access to those applications. Any database can be negatively impacted by data access bottlenecks caused by I/O wait, which can also result in slow application response time. I/O wait is the total duration that the working processes of a database or application session are blocked while waiting for I/O operations to complete. High I/O wait can mean that your storage device is providing “unacceptable” services to your database or other mission critical applications, resulting in low customer satisfaction and employee productivity. High latency or low throughput can result with significant I/O Wait, resulting in poor database response times for end users.

The following assessments of Oracle and MS SQL Server performance shows that by eliminating I/O bottlenecks, database performance can realize an up to 11-fold improvement.

Bookmark/FavoritesFacebookDeliciousDiggGoogle+LinkedInTechnorati FavoritesStumbleUponTumblrTwitterShare
Posted January 12th, 2011
Gareth Taube

Digital Trowel Cuts Web Mining Processing Time in Half

By Gareth Taube, Vice President Marketing, Kaminario
Bookmark/FavoritesFacebookDeliciousDiggGoogle+LinkedInTechnorati FavoritesStumbleUponTumblrTwitterShare

Digital Trowel, a leading corporation in Web mining, collects, distills and disseminates data from the Internet to provide the knowledge needed to enable its customers to make smart business decisions. It harvests data (structured and unstructured) through Web crawlers and maintains multiple databases with a combined total of more than 10 billion records. Digital Trowel’s applications include data discovery and aggregation of business intelligence information, sentiment analysis and market mapping, and brand awareness analysis.

Learn how they cut their Web mining processing time in half.

Read the full Digital Trowel case study.

Bookmark/FavoritesFacebookDeliciousDiggGoogle+LinkedInTechnorati FavoritesStumbleUponTumblrTwitterShare
Posted January 5th, 2011
Eyal Markovich

Replication in a High Performance Environment

By Eyal Markovich
Bookmark/FavoritesFacebookDeliciousDiggGoogle+LinkedInTechnorati FavoritesStumbleUponTumblrTwitterShare

In many industries, the replication of a company’s storage is a crucial operational task. It may even define a company’s ability to survive major events. In many cases, replicating data could also be a government regulated action. When it comes to implementing replication for high performance storage, one should take into consideration the many characteristics of replication and the storage it is employed on. The Kaminario whitepaper on Replication in a High Performance Environment describes what replication is, followed by a discussion on how to implement replication on high performance storage.

Bookmark/FavoritesFacebookDeliciousDiggGoogle+LinkedInTechnorati FavoritesStumbleUponTumblrTwitterShare
Posted January 3rd, 2011
Eyal Markovich

The Importance of Write Performance to RDBMS Applications

By Eyal Markovich
Bookmark/FavoritesFacebookDeliciousDiggGoogle+LinkedInTechnorati FavoritesStumbleUponTumblrTwitterShare

Both Oracle and SQL Server perform writes that are asynchronous to the user sessions (not to be confused with asynchronous I/O, which is something else). This means that when a user session modifies rows in a table, the modified pages will be physically written to the table at a later stage and will not add additional time to the DML operations.

So, we might say that we don’t care much about storage write performance in Oracle and SQL. Correct?

In short – the answer is no. Read the rest of this entry »

Bookmark/FavoritesFacebookDeliciousDiggGoogle+LinkedInTechnorati FavoritesStumbleUponTumblrTwitterShare