lat_rpc.8 1.6 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748
  1. .\" $Id$
  2. .TH LAT_RPC 8 "$Date$" "(c)1994 Larry McVoy" "LMBENCH"
  3. .SH NAME
  4. lat_rpc \- measure interprocess communication latency via Sun RPC
  5. .SH SYNOPSIS
  6. .B lat_rpc
  7. .I -s
  8. .sp .5
  9. .B lat_rpc
  10. .I hostname [udp|tcp]
  11. .sp .5
  12. .B lat_rpc
  13. .I -hostname
  14. .SH DESCRIPTION
  15. .B lat_rpc
  16. is a client/server program that measures interprocess
  17. communication latencies. The benchmark passes a token back and forth between
  18. the two processes (this sort of benchmark is frequently referred to as a
  19. ``hot potato'' benchmark). No other work is done in the processes.
  20. .PP
  21. This benchmark may be compared to the TCP and UDP forms of the same benchmark
  22. to accurately see the cost of using RPC versus the cost of using plain
  23. old TCP or UDP sockets. It is worth noting that the RPC form is passing
  24. back and forth a single byte, not some long complicated record.
  25. .PP
  26. .B lat_rpc
  27. has three forms of usage: as a server (-s), as a client (lat_rpc localhost), and
  28. as a shutdown (lat_rpc -localhost).
  29. .P
  30. The client form may specify the protocol over which the RPCs are performed.
  31. The default is ``udp''.
  32. .SH OUTPUT
  33. The reported time is in microseconds per round trip and includes the total
  34. time, i.e., the context switching overhead is included.
  35. Output format is like so
  36. .sp
  37. .ft CB
  38. RPC/udp latency using localhost: 1344 microseconds
  39. .br
  40. RPC/tcp latency using localhost: 2089 microseconds
  41. .ft
  42. .SH ACKNOWLEDGEMENT
  43. Funding for the development of
  44. this tool was provided by Sun Microsystems Computer Corporation.
  45. .SH "SEE ALSO"
  46. lmbench(8), svcudp_create(3), svctcp_create(3), svc_register(3),
  47. svc_run(3), svc_sendreply(3), svc_freeargs(3), clnt_create(3),
  48. clnt_control(3), clnt_call(3).