= "\
hpcserver is an optional component that works with hpctraceviewer to enable\n\
visualizations of larger databases. The disk access and heavy processing\n\
required by hpctraceviewer are offloaded to hpctraceserver, which can run in\n\
parallel with MPI. The results are streamed back to a connected hpctraceviewer\n\
client in a way that permits the same interactivity as running hpctraceviewer\n\
with the data located locally.\n\
\n\
Unlike standard web servers, for example, hpcserver is designed to run only\n\
while in use, and not perpetually as a daemon on background process. hpcserver\n\
also runs with the permissions of the user that launched it and allows the\n\
hpctraceviewer client to specify and access any database that hpcserver can read.\
\n\
Options: General\n\
-V, --version Print version information.\n\
-h, --help Print this help.\n\
-c, --compression Enables or disables compression (on by default)\n\
Allowed values: on off \n\
-p, --port Sets the main communication port (default is 21590)\n\
Specifying 0 indicates that an open port should be \n\
chosen automatically.\n\
-x, --xmlport Sets the port on which the experiment.xml file will be\n\
transmitted. Specifying 0 or not including this flag\n\
indicates that the port will be auto-negotiated with\n\
the client. Specifying 1 indicates that the xml will\n\
be transferred on the main data port.\n\
\n\
"