My App Server Is Slow! What Should I Do?

If you’re a SaaS provider, the last thing you want your customers to see when they use your service is “Sorry, our systems are down. Try back later.”

Systems do go down and you have to figure out what the problem is and get your service available to your customers ASAP.

But troubleshooting network application issues can be like searching for a needle in three haystacks.

Wanted: Experienced Network Application Troubleshooters

Where do you start?

Do you check network logs? What if you find that something is using all your bandwidth? What if some process has your CPUs pegged? What if all the memory on a box is being used?

Is a network component going bad? Is there a problem with your application server? Is there a record lock in a key table at the database level? Or is it something else?

When Reboot Is Not the Needle You Are Looking For

The software company in this story was getting reports that it’s users were encountering gradually worsening slowdowns when attempting to access information from the web applications. Their IT team tried to proactively trouble shoot the problem in the best way they knew how. They rebooted.  But rebooting server and network equipment only temporarily relieved the issues and customers eventually reported more slowdowns.

The software organization was focused on servicing their users and addressing issues related to the software, not troubleshooting a slow system. Their clients expected instant access to online database service and slow connections would not be tolerated for long.

When Your Internal IT Needs a Detective on the Case

When it became clear that the ability to find the needle with the right manpower fell outside of the programmers’ IT skill set, the software company turned to Link High Technology, an outside IT provider able to provide both the tools and expertise to troubleshoot and remediate the issue.

Link High quickly recognized that network monitoring alone would not be able to detect the issue. A team of certified technicians, or “detectives” (of sorts), were brought on to augment the software company’s internal team to isolate the issue. Utilizing the finite details of the detected errors on the server, Link High’s team was able to monitor the errors and detect what was happening to the requests for data passing through multiple web servers.

Finding the Needle (In 3 Haystacks)

Finding the needle was an extensive process of testing and procedures you would expect to receive from a highly qualified Managed Service Provider like Link High. The needle was identified through a series of tests:

  • Link High’s team set up and monitored traffic in 5-hour intervals, paying close attention to peak-usage during working hours.
  • They noted that networking switches reached an optimal amount of bandwidth saturation. Realizing this was not the problem, they kept digging.
  • Due to the severity of client-reported issues, Link High switched focus to both the internal database servers and an internal web server, a total of three locations. Where many IT departments would probably stop here, Link High knew the problem had to go beyond bandwidth and performance issues and focused on individual interface traffic flows, or, to put it simply, how did the server relay info to the databases, especially at mid-point junctures.
  • The Link High team’s detective work performed tests to see if the problem was networking equipment, servers, or the databases that the information is housed in.

After thorough analysis of data flow from customer to our client’s servers, the Link High team localized the problem to a particular server’s response time. The initial request for information going to the databases was processed at normal speeds, but it was the delivery of information back to the user that was the holdup. It turned out to be the mid-point server that was bogging down data flow and not the databases housing the information. The slow down at this mid-point junction was causing a major speed bump.

Once the problem was found, the Link High experts moved onto remediation.

  • The first order of business was to update the switch firmware so that even higher bandwidth could be optimized.
  • Code in the older internal server was also updated to make sure the proper communication occurred between incoming and outgoing data requests.
  • This code maintenance reduced queues for information from stacking up and keeps the entire network (from reception to back-office coders) up and operational.

How Would a Managed Approach Have Prevented This Problem Before It Began

Programs, like the one deployed by Link High, are designed to monitor and detect either the most common or unique IT situations for a specific problem. And, as in this case, experts like Link High have the tools and expertise that an internal IT team doesn’t. Managed Service Providers like Link High are often called in to help internal IT teams solve challenging problems or address rare issues that IT teams only see occasionally. As a professional, certified IT provider, Link High and other managed service providers work on a broad range of IT challenges.

If you or your team is experiencing unexpected or unexplained lapses in service, Link High has dedicated project managers and engineers that have the experience to analyze and mitigate performance issues. We work closely with many internal IT teams as well as being the sole managed IT provider for businesses. With a dogged attention to detail to remedy performance issues, we analyze, isolate the problem, and get things running smoothly again so that you can provide your customers the services they depend on.

Download the e-book here

Leave a Comment

Get In Touch

We're not around right now. But you can send us an email and we'll get back to you, asap.

Not readable? Change text.