flow/packet loss through L3 C3560, pings OK
Bookmark this page:
 Yahoo!
 Google
 Windows Live
 del.icio.us
 digg
 Netscape
|
|
||||||||||||||||
|
Posted by Phil Begriffenfeldt on May 1, 2008, 12:59 am
Please log in for more thread options default gateway for a web cluster, which was doing about 120mbps of traffic, 5kpps each in and out. However, users noticed slow page loads, broken inline images, etc. I was able to ping all the servers from outside the 3560G with zero packet loss in tens of thousands of 1500-byte pings. I moved the web cluster to a C6509 (same interface config) and the issue disappeared. Web client experience was noticably impacted, so if it were simple packet loss, I think I would have seen it with ping. It seemed as though the issue was related either to the type of traffic (plain http) or flow (lots of flows). The 3560 has a pretty vanilla config; the web cluster traffic was being routed between a "no switchport" interface and a Vlan interface. I did notice that the "no switchport" interface had "ip route-cache same-interface" configured, and I'm not sure why. Also, the 3560 is carrying about 7k external routes, but I monitor it to make sure it doesn't hit the limit. I didn't see any clues in syslog. Phil | ||||||||||||||||
|
Posted by stephen on May 1, 2008, 5:37 pm
Please log in for more thread options we had some issues with the 10/100 versions with buffer tuning where we had problems with traffic bursts overwhelming the buffers, esp when you turn QoS on as you effectively reduce the buffer pool for any 1 QoS type by 75%. If you have several GigE connected servers contending for a congested or rate limited port this could be an issue. there are some commands to look at the buffers - something like show platform port-asic statistics..... you want the drop stats for any overloaded outbound ports. >
> Web client experience was noticably impacted, so if it were simple > packet loss, I think I would have seen it with ping. It seemed as > though the issue was related either to the type of traffic (plain http) > or flow (lots of flows). > > The 3560 has a pretty vanilla config; the web cluster traffic was being > routed between a "no switchport" interface and a Vlan interface. I did > notice that the "no switchport" interface had "ip route-cache > same-interface" configured, and I'm not sure why. Also, the 3560 is > carrying about 7k external routes, but I monitor it to make sure it > doesn't hit the limit. I didn't see any clues in syslog. you need "sdm prefer routing" in the config to handle lots of IP routes - if not they overflow the hardware forwarding table and get dealt with in software. >
--
> Phil Regards stephen_hope@xyzworld.com - replace xyz with ntl | ||||||||||||||||
|
Posted by Phil Begriffenfeldt on May 2, 2008, 9:28 am
Please log in for more thread options stephen wrote:
> you need "sdm prefer routing" in the config to handle lots of IP routes
Thanks for reminding me. I did set that last May (it's logged), and then power-cycled the switch, but I do not appear to have verified "show sdm" after the power cycle. Now I see that the switch is using default/desktop, which could be the source of my trouble. Weird. | ||||||||||||||||
|
Posted by stephen on May 2, 2008, 3:11 pm
Please log in for more thread options > stephen wrote:
> > you need "sdm prefer routing" in the config to handle lots of IP routes
>
> Thanks for reminding me. I did set that last May (it's logged), and > then power-cycled the switch, but I do not appear to have verified "show > sdm" after the power cycle. Now I see that the switch is using > default/desktop, which could be the source of my trouble. Weird. yes - hardware forwarding tables will fill with 1 to 2k routes. everything that arrives after the tables fill goes in software forwarding - so whether it is an irritation or a disaster depends on the order the routes arrive. Not a fun thing to trouble shoot, but it does log an "out of space" message - shame Cisco couldnt make it obvious what it is an error about.... -- Regards stephen_hope@xyzworld.com - replace xyz with ntl | ||||||||||||||||
| Similar Threads | Posted |
| flow/packet loss through L3 C3560, pings OK | May 1, 2008, 12:59 am |
| Files on a C3560? | March 1, 2006, 6:48 am |
| C3560 vs C3750 switches | July 13, 2005, 3:31 pm |
| VLANs routed with C3560 and Proxy ARP | August 10, 2008, 1:08 pm |
| I have strange routing problem with my C3560 switch. | March 7, 2007, 7:20 am |
| Wireless pings | June 3, 2008, 4:39 pm |
| Cisco 800 - only repond to pings from certain addresses? | May 5, 2006, 3:17 am |
| ACL: Does "permit IP" allow ICMP traffic like pings? | January 4, 2007, 12:01 pm |
| Re: Linksys wireless-G notebook adapter pings Cisco web each second | March 23, 2008, 11:17 pm |
| Re: Linksys wireless-G notebook adapter pings Cisco each second | March 24, 2008, 1:57 am |
| At a loss | October 5, 2005, 6:30 am |
| Loss of VPN Access Using Pix 501 | July 21, 2006, 10:19 pm |
| 50% packet loss ? | February 25, 2008, 1:00 pm |
| some commands about loss packet | February 13, 2006, 8:33 am |
| WAN packet loss/latency | April 24, 2006, 9:23 pm |
> default gateway for a web cluster, which was doing about 120mbps of
> traffic, 5kpps each in and out. However, users noticed slow page loads,
> broken inline images, etc.
>
> I was able to ping all the servers from outside the 3560G with zero
> packet loss in tens of thousands of 1500-byte pings. I moved the web
> cluster to a C6509 (same interface config) and the issue disappeared.