Open GoogleCodeExporter opened 8 years ago
Ok, so when an iperf3 client tries to connect to an iperf2 server, the client
loops forever in iperf_run_client() waiting for a response. I need to do a
little more investigation to figure out if it's possible/reasonable to add a
check in that loop for iperf2.
Original comment by jef.posk...@gmail.com
on 3 Jun 2013 at 12:54
Hard to do. Will try in future release.
Original comment by bltier...@es.net
on 23 Jul 2013 at 5:35
Original comment by bltier...@es.net
on 18 Dec 2013 at 9:17
Original comment by bltier...@es.net
on 18 Dec 2013 at 10:40
Iperf3 and iperf 2 works in a diff way and iperf3 control socket does not have TMO it just hangs on select waiting to for some response.
socket(PF_INET, SOCK_STREAM, IPPROTO_IP) = 3 connect(3, {sa_family=AF_INET, sin_port=htons(5001), sin_addr=inet_addr("127.0.0.1")}, 16) = 0 write(3, "XXX.1387736273.832197.4c21154a48"..., 37) = 37 clock_gettime(CLOCK_PROCESS_CPUTIME_ID, {0, 1689286}) = 0 getrusage(RUSAGE_SELF, {ru_utime={0, 0}, ru_stime={0, 1000}, ...}) = 0 select(4, [3], [], NULL, NULL <================ Iperf3 client hangs here ..
Well to handle this scenario code also changes required for iperf2 which I dont think really possible or is it ? This is not something we can control . I do have some approches
1. Control socket Setting non-block
2. Control socket TMO
Original comment by susant%redhat.com@gtempaccount.com
on 22 Dec 2013 at 6:27
Noting that iperf2 and iperf3 have different default ports (5001 vs. 5021), I
think that this problem won't come up real often.
Original comment by bmah@es.net
on 3 Feb 2014 at 10:50
In that case it should be doable to check if 5001 is open/iperf2 when iperf3
gets a 'connection refused' on 5021.
I also (naively) assumed that iperf3 would be able to connect to an iperf2
server, all i got was:
iperf3: error - unable to connect to server: Connection refused
would be helpful if it would mention the incompatibility.
Original comment by goo...@vanbragt.com
on 12 Feb 2015 at 5:12
Original issue reported on code.google.com by
bltier...@es.net
on 30 May 2013 at 11:21