Understanding the System

A typical DAM setup consists of end users accessing DAM via a load balancer. The DAM instance might be part of a clustered setup, where each DAM instance runs in a Java™ Virtual machine process on either a physical machine or a virtual machine. DAM storage is either provided by a RAID disk if there are single-machine setups, or a managed network-attached storage if there are clustered setups.

The following legend describes the possible performance pitfall areas with some solutions, as appropriate.

Network connection to end user A slow network connection can cause throughput issues, and in some rare cases also latency issues. Sometimes the user has a slow connection from the ISP, especially in intranets. This is a sign of incorrect network topology.

Temporary File System A slow local file system can cause interactive performance problems, especially when searching, because the search indexes are stored on the local disk. It can also cause asset processing problems if the command-line process is being used.

AEM DAM Finder Interactive performance problems, often experienced in searches are caused by high CPU utilization due to many concurrent users or other CPU-consuming processes on the same instance. Moving from virtual machines to dedicated machines and making sure no other services run on the machine can help improving performance. If high CPU load is caused due to asset processing and many concurrent users, Day recommends adding additional cluster nodes.

AEM DAM Workflow Long-running workflow processes during asset ingestion cause asset processing performance problems. Depending on the type of assets being processed, this can indicate CPU over-utilization. Day recommends that you reduce the number of other processes running on the system and to increase the number of available CPUs by adding cluster nodes.

NAS Connectivity Poor network connectivity to the NAS causes interactive performance problems, because accessing new nodes during asset processing is slowed down due to network latency. Also, slow network throughput adversely affects throughput, but also asset processing performance, because loading and saving renditions is slowed down.

Reasons for bad latency and throughput in a NAS are network topology or NAS over-utilization by other services.

Network Attached Storage Over-used network-attached storage systems can cause an array of problems:

  • Low disk space is a frequently encountered trouble that can be prevented through proper sizing of a DAM project.
  • High disk latency propagates into slow access times for CRX and may result in interactive performance problems.
  • Low disk throughput may result in low performance for CQ5 DAM.