Jump to content

Solutions

Technology - Database

Diagnose Database Performance Problems in Five Minutes Flat

Many companies rely so heavily on the functions of databases that normal business operations cannot continue if databases are unavailable. Databases represent a fair amount of management complexity; coupled with interactions with the application, network, and storage layers, it’s no wonder databases often are blamed for performance issues. Some database-related performance challenges include:

  • Lack of visibility into production systems – Various tools offer profiling of database performance during development, testing, or staging. However, due to the high overhead imposed by these monitoring tools, they rarely can be used in production. This shortcoming leads to a severe lack of visibility into production databases.
  • Cross-domain issues that impact the database – Similar to the network, the database gets more than its share of finger pointing in a blamestorm session. The application, network, and database layers are so interconnected that it’s often difficult to isolate the root cause.
  • Database as a shared resource – In this scenario, the IT group manages a set of shared databases that support many applications. When a single application applies too much load on the shared database, performance could slow down across the board, making it very difficult to find the true culprit.
  • Mistaking performance issues for capacity problems – When database transactions start to slow, a common knee-jerk reaction is to buy more hardware. However, without understanding the underlying issues, such as transaction timing or monopolization of resources, buying more capacity does not necessarily solve the problem.

With the ExtraHop system, you can pinpoint database issues in just minutes. The ExtraHop system can help.

  • Tune database performance – The ExtraHop Application Delivery Assurance system provides a play-by-play account of database performance that includes detailed transaction-performance metrics broken down by method, table, and stored procedure. At a glance, you can see who is calling any given method, how frequently, and which specific operations are performed. This arms you with the information you need to make optimal tuning decisions.
  • Troubleshoot database problems – When it comes to troubleshooting, the ExtraHop system provides a full who, what, when, where, and why breakdown for most issues. At the click of a button, you quickly can find out who is experiencing errors, what are the error details, when did they occur, and which tables in which databases are impacted.
  • Facilitate capacity planning – With detailed transaction timing and trends, you can see when more database resources are needed to maintain adequate performance as well as any throttling that might be occurring. The ExtraHop system provides metrics all the way up and down the protocol stack to highlight any network- and application-specific issues.

Customer Vignette

Who: Online Media Company.

Challenge: New application rollout leading to perceived database-performance issue.

The ExtraHop System in Action: The ExtraHop Application Delivery Assurance system was monitoring the application rollout and quickly found that the number of database errors jumped from 2–3 per second to 2000+ per second. With the detailed error and method information, the customer was able to trace this problem to an application error and exonerated the database. A quick patch from the application team alleviated the issue entirely.

ROI: 99% reduction in database errors and intermittent errors.