Re: MICE-DISCUSS Digest - 1 Oct 2011 to 3 Oct 2011 (#2011-99)
I've seen that when flood-pinging BSD boxes (but not Linux). If you can throttle the rate down (-i 0.1), rather than flood pinging, then the artifact may go away.
Sending 1000, 100-byte ICMP Echos to 69.147.218.1, timeout is 2 seconds: !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!.!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!.!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!.!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
######################################################################## To unsubscribe from the MICE-DISCUSS list, click the following link: http://lists.iphouse.net/cgi-bin/wa?SUBED1=MICE-DISCUSS&A=1
I do believe that it is something in the OS of the route server. Unfortunately in the Cisco IOS, I can not adjust the rate, however I found that it always drops the 200'th packet. If I do multiple individual pings back to back (with one second or more inbetween them) that repeat 199 times, then there is never any loss. It also explains why I can do the same test to other devices across my MICE connection, and do not drop every 200th packet. Also by being specific to the 200'th packet in a row that tells me that there is not a speed/duplex mismatch, or a cable issue as I had originally worried about (since they would be random, or on a specific time interval, not based on a exact packet number). Jeremy -----Original Message----- From: MICE Discuss [mailto:MICE-DISCUSS@LISTS.IPHOUSE.NET] On Behalf Of Bradley Urberg Carlson Sent: Tuesday, October 04, 2011 9:46 AM To: MICE-DISCUSS@LISTS.IPHOUSE.NET Subject: Re: [MICE-DISCUSS] MICE-DISCUSS Digest - 1 Oct 2011 to 3 Oct 2011 (#2011-99) I've seen that when flood-pinging BSD boxes (but not Linux). If you can throttle the rate down (-i 0.1), rather than flood pinging, then the artifact may go away.
Sending 1000, 100-byte ICMP Echos to 69.147.218.1, timeout is 2 seconds: !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!.!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!.!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!.!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
######################################################################## To unsubscribe from the MICE-DISCUSS list, click the following link: http://lists.iphouse.net/cgi-bin/wa?SUBED1=MICE-DISCUSS&A=1 ######################################################################## To unsubscribe from the MICE-DISCUSS list, click the following link: http://lists.iphouse.net/cgi-bin/wa?SUBED1=MICE-DISCUSS&A=1
Sure enough - 200 responses per second, good troubleshooting guys! [micemn-01 /etc]$ sysctl -d net.inet.icmp.icmplim net.inet.icmp.icmplim: Maximum number of ICMP responses per second [micemn-01 /etc]$ [micemn-01 /etc]$ sysctl net.inet.icmp.icmplim net.inet.icmp.icmplim: 200 [micemn-01 /etc]$ -----Original Message----- From: MICE Discuss [mailto:MICE-DISCUSS@LISTS.IPHOUSE.NET] On Behalf Of Jeremy Lumby Sent: Tuesday, October 04, 2011 10:27 AM To: MICE-DISCUSS@LISTS.IPHOUSE.NET Subject: Re: [MICE-DISCUSS] MICE-DISCUSS Digest - 1 Oct 2011 to 3 Oct 2011 (#2011-99) I do believe that it is something in the OS of the route server. Unfortunately in the Cisco IOS, I can not adjust the rate, however I found that it always drops the 200'th packet. If I do multiple individual pings back to back (with one second or more inbetween them) that repeat 199 times, then there is never any loss. It also explains why I can do the same test to other devices across my MICE connection, and do not drop every 200th packet. Also by being specific to the 200'th packet in a row that tells me that there is not a speed/duplex mismatch, or a cable issue as I had originally worried about (since they would be random, or on a specific time interval, not based on a exact packet number). Jeremy -----Original Message----- From: MICE Discuss [mailto:MICE-DISCUSS@LISTS.IPHOUSE.NET] On Behalf Of Bradley Urberg Carlson Sent: Tuesday, October 04, 2011 9:46 AM To: MICE-DISCUSS@LISTS.IPHOUSE.NET Subject: Re: [MICE-DISCUSS] MICE-DISCUSS Digest - 1 Oct 2011 to 3 Oct 2011 (#2011-99) I've seen that when flood-pinging BSD boxes (but not Linux). If you can throttle the rate down (-i 0.1), rather than flood pinging, then the artifact may go away.
Sending 1000, 100-byte ICMP Echos to 69.147.218.1, timeout is 2 seconds: !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!.!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!.!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!.!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
######################################################################## To unsubscribe from the MICE-DISCUSS list, click the following link: http://lists.iphouse.net/cgi-bin/wa?SUBED1=MICE-DISCUSS&A=1 ######################################################################## To unsubscribe from the MICE-DISCUSS list, click the following link: http://lists.iphouse.net/cgi-bin/wa?SUBED1=MICE-DISCUSS&A=1 ######################################################################## To unsubscribe from the MICE-DISCUSS list, click the following link: http://lists.iphouse.net/cgi-bin/wa?SUBED1=MICE-DISCUSS&A=1
participants (3)
-
Anthony Anderberg -
Bradley Urberg Carlson -
Jeremy Lumby