Don’t Forget Your System Calls

DLT’s Greg Agana provides a friendly reminder for everyone out there that troubleshooting difficult software application issues doesn't have to be as hard as quantum mechanics.

When you run into an issue, like you would in Solaris or RHEL, there are a few steps we go through to figure out what exactly is causing the problem:

  1. Check Installation Process
  2. Check Configuration
  3. Check Application Logs                                                     
  4. Check Operating System Logs (Event Viewer, etc)           
  5. Double-check Installation Process
  6. Double-check Configuration
  7. Pull Your Hair Out (HairClub and Bosley)
  8. Sniff Packets (snoop, tcpdump, and wireshark)

Sometimes you just can’t find the smoking gun.  The rest of your systems are supposedly the same.  You even go as far as going to an older version of the software, but the results are the same.  What is different?  What changed?

Tools that we often forget about are our system’s call tracing tools, like strace and truss.  These tools are invaluable to the troubleshooting of incredibly difficult issues. When other troubleshooting tools and processes are just not cutting it, we utilize these tools all the time in the DLT Service Center.  We are able to capture great information that assists us in the absolute fix of the underlying issue. Anything else we can add to this piece?

Microsoft Truss

It been brought to my attention recently that there is a truss for Windows.  I’ve yet to come across it, so I don’t have much to offer on how well they work in helping solve However, since information exchange is how we collectively grow our knowledge base, I’d love to hear from you If you have used the Microsoft truss.

In the mean time, if you think you have hit the end of the road in your troubleshooting process, I hope you remember to trace your system calls.

Until next time…