Most of us have experienced this at one time or another… A critical application on the network is operating at a painfully slow speed. End-user tension increases as “everything bogs down.” People begin to point fingers as support engineers scramble to find a solution…
There are a variety reasons that an application can suffer a decrease of speed on the network.
Initially, many people may assume that the application itself is to blame for slow performance, but typically this is not the case. More often than not, the issue is related to other factors.
When confronted with an application that is running slowly on the network, two simple steps can be used to isolate and solve the problem:
- Find the bottlenecks by analyzing and managing your application - to find the root cause of slowness on the network, isolate the problem by asking the following questions:
- What part of the Network is slow? (LAN, Campus, WAN, Internet?)
- What is the bandwidth usage for each direction (in and out)?
- What type of data (protocol) and application (ports) are using most of the bandwidth?
For example, moving high resolution images on a picture archiving and communication system (PACS) can dramatically slow a network. Bottlenecks can be found by monitoring the application port that generates the traffic, identifying the source node and focusing on the affected subnet.
- Consider automation strategies to accelerate and improve service to your customer - some network delays can be the result of a lack of resources on the OS or storage level. Here are some items you should look into while monitoring resources:
- What is CPU and memory usage on operating systems used to host network applications? By monitoring the CPU and memory usage on the application level, you can easily pinpoint application bottlenecks.
- Do any of the applications run on virtual machines (VMs)? What resources are allocated to VMs?
- How is Storage Area Networking (SAN) or Network Attached Storage (NAS) being used? What are the IOPS and the read/write transfer rates?
Another example from healthcare - many picture archiving and communication systems (PACS) are using SAN for storage. Some storage may be over-used while other storage may not be allocated at all. Monitoring of application I/O to the database (looking into the application log files or database metrics, etc.) can help with the automation of the troubleshooting process.
Speeding up a slow application on the network is easy to do when you know what to look for. We hope that these tips will help you to provide a better user experience to your customers.
Author: Nir Yosha, Pre Sales Manager - NextNine Inc.