A number of tools and techniques are available to monitor the operation of the server, its environment, and applications.
Windows platforms provide a number of diagnostic tools:
Other Windows tools are described under “Runtime monitoring tools”.
UNIX systems also provide a number of tools:
On HP-UX, numerous commands and tools are available including:
tusc – to list all system calls; useful for tracing application crashes.
vmstat – for virtual memory statistics.
top – for a list of processes using the most CPU.
On AIX, use the System Admin tools.
On Solaris, numerous commands are available, including:
top – to see which processes are consuming the most CPU.
vmstat – displays CPU and memory usage.
pstack – for stack dump analysis.
pmap – lists which libraries are loaded and their load locations.
pldd – similar to pmap but includes less information.
pfiles – lists file descriptors, sockets, files, and so on.
truss – displays what is happening with file descriptors, signals, O/S interactions with a process, and so on. truss is useful if you know that a specific action leads to a specific problem.
truss can generate large volumes
of output. To reduce truss output, use truss
-p.
