Skip to Content.
Sympa Menu

perfsonar-user - Re: [perfsonar-user] a nit on owping timestamp output

Subject: perfSONAR User Q&A and Other Discussion

List archive

Re: [perfsonar-user] a nit on owping timestamp output


Chronological Thread 
  • From: Aaron Brown <>
  • To: Alan Whinery <>
  • Cc: " Users" <>
  • Subject: Re: [perfsonar-user] a nit on owping timestamp output
  • Date: Mon, 9 Jun 2014 19:39:09 +0000
  • Accept-language: en-US

Hey Alan,

It’d probably be reasonable to add a flag to bwctl (and if possible the tools it calls) to let the user configure the timestamp output. I’m not positive if we’d want to go full-on timezone, but maybe something like “—gmtime”, and have a flag in the protocol to say “i want all timestamps to be output in gmtime” or something similar. I’m not sure if any of the other tools spit out a timestamp (iperf3 might) so it may be a more “one-use” flag, but I could see a use for it.

Cheers,
Aaron

On Jun 5, 2014, at 4:00 PM, Alan Whinery <> wrote:

I notice that owping returns the "first" and "last" timestamps as a zone-less time reference, apparently using local time on the sender as the reference.  This matters little when one is using the command line and getting a report for each of two tests, one each direction, in a terminal. A newer, emerging use-case, with one-way tests invoked through bwctl, possibly from a third host, possibly saving output to parse, can leave one with un-clear timestamps.

Below is an example, with "bwping -T owamp", invoking tests between two bwctl hosts.  I just stumbled onto this because Ubuntu-armhf installed with local=GMT, and Raspbian installs with local=HST.  The tests were run one right after the other, but appear to have occurred 10 hours apart.

The obvious fix would be to simply have owping always get time with gmtime() versus localtime() (or similar, I haven't looked at code). Of course, one would need to consider breaking existing owping-output-parsers.  Using localtime even be a flaw/oversight in owamp implementation, since it seems counter-intuitive for such things.

-Alan

root@armv6l-b8-27-eb-c1-45-11:/etc/owampd#  bwping -T owamp -N 50 -i 0.1 -s hostA.hawaii.edu A AESKEY bob /etc/bwctld/bwctld.keys -c hostB.hawaii.edu A AESKEY bob /etc/bwctld/bwctld.keys
bwping: 15 seconds until test results available

SENDER START
bwctl: start_endpoint: 3610982384.435748
bwctl: run_endpoint: receiver: hostB.hawaii.edu
bwctl: run_endpoint: sender: hostA.hawaii.edu
bwctl: exec_line: owping -t -c 50 -i 0.100000 [hostB.hawaii.edu]:4000
bwctl: run_tool: tester: owamp
bwctl: run_tool: receiver: hostB.hawaii.edu
bwctl: run_tool: sender: hostA.hawaii.edu
bwctl: start_tool: 3610982390.995661
Approximately 8.1 seconds until results available

--- owping statistics from [hostA.hawaii.edu]:39037 to [hostB.hawaii.edu]:45477 ---
SID:    80ab5714d73b37f719eed0d875a81922
first:  2014-06-05T18:39:52.181
last:   2014-06-05T18:39:57.101
50 sent, 0 lost (0.000%), 0 duplicates
one-way delay min/median/max = 0.196/0.3/0.462 ms, (err=0.414 ms)
one-way jitter = 0.1 ms (P95-P50)
Hops = 5 (consistently)
no reordering

bwctl: stop_tool: 3610982400.301793
bwctl: stop_endpoint: 3610982401.804053

SENDER END
root@armv6l-b8-27-eb-c1-45-11:/etc/owampd#  bwping -T owamp -N 50 -i 0.1 -c hostA.hawaii.edu A AESKEY bob /etc/bwctld/bwctld.keys -s hostB.hawaii.edu A AESKEY bob /etc/bwctld/bwctld.keys
bwping: 15 seconds until test results available

SENDER START
bwctl: start_endpoint: 3610982423.024834
bwctl: run_endpoint: receiver: hostA.hawaii.edu
bwctl: run_endpoint: sender: hostB.hawaii.edu
bwctl: exec_line: owping -t -c 50 -i 0.100000 [hostA.hawaii.edu]:4000
bwctl: run_tool: tester: owamp
bwctl: run_tool: receiver: hostA.hawaii.edu
bwctl: run_tool: sender: hostB.hawaii.edu
bwctl: start_tool: 3610982429.660192
Approximately 8.1 seconds until results available

--- owping statistics from [hostB.hawaii.edu]:51330 to [hostA.hawaii.edu]:54300 ---
SID:    80abd6f4d73b381dc3ef584157befacb
first:  2014-06-05T08:40:30.808
last:   2014-06-05T08:40:34.978
50 sent, 0 lost (0.000%), 0 duplicates
one-way delay min/median/max = 0.721/0.9/1.02 ms, (err=0.414 ms)
one-way jitter = 0.1 ms (P95-P50)
Hops = 6 (consistently)
no reordering

bwctl: stop_tool: 3610982437.321193
bwctl: stop_endpoint: 3610982439.333351

SENDER END





Archive powered by MHonArc 2.6.16.

Top of Page