2015-08-04 20:43 GMT+08:00 Alexey Kodanev <alexey.koda...@oracle.com>: > Hi, > On 08/03/2015 05:59 PM, Cyril Hrubis wrote: >> >> Hi! >>> >>> Some switches need a while to learn the new assigned address, which make >>> the >>> connectivity check failed. Make PING_MAX adjustable to let user can >>> pre-define >>> it based on their own environment. >> >> I still think that the less parameters user has to set before the >> testing the better. Alexey what do you think about this? >> > > May be we should add this connectivity checking to test_net.sh as a function > that can check both ICMPv4/6. It shouldn't require any additional parameters > and max ping time could be set to some reasonable value. I don't mind > increasing the time because it returns immediately after successful reply; I > just want to understand why 10 seconds not enough and what time is enough > for those switches, for v4 and v6 in particular?
In our lab, we use cisco switches and it take about 30s to learn the new manually assigned address for both v4 and v6. And some times we run the test between guests via linux birdge, then we can ping remote directly after assign the address. That's why I would like to make PING_MAX adjustable, it may take too much time if we hard code a large value if run tests between guests. Thanks Hangbin > > Unfortunately, not all tests cleaned up in stress/ and we would need to keep > the old version as well. tcp/ udp/ route/ multicast/ not include test_net.sh > yet. > > Thanks, > Alexey > ------------------------------------------------------------------------------ _______________________________________________ Ltp-list mailing list Ltp-list@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/ltp-list
