This was fixed with one simple command: qos queue-softmax-multiplier 1200 Under 99% load we'd see 10-50% output drops. If I keep them at 99% utilization, I get up to 0.01% output drops.Īnother issue I had was crazy-high drops from our 3850 to the 2960s. Received 1671991323 broadcasts (1095981898 multicasts)Ġ input errors, 0 CRC, 0 frame, 0 overrun, 0 ignoredĠ watchdog, 1095981898 multicast, 0 pause inputĠ input packets with dribble condition detectedġ57866660989 packets output, 206736907893497 bytes, 0 underrunsĠ output errors, 0 collisions, 1 interface resetsĠ lost carrier, 0 no carrier, 0 pause outputĠ output buffer failures, 0 output buffers swapped outĮdit OK, the 2960 to 2960 packet drops are pretty low. Last clearing of "show interface" counters never Input flow-control is off, output flow-control is unsupported Hardware is Gigabit Ethernet, address is (bia ) GigabitEthernet1/0/48 is up, line protocol is up (connected)
#Add helper app to intermapper full
This is what the full output looks like on an interface from a "sending" switch: # show interfaces GigabitEthernet 1/0/48 InterMapper calls these discards "Link Errors", and keeps notifying me that my switches are accumulating errors like crazy.Īre these discards normal? Safe to ignore? I've been trying to diagnose performance issues on a 3850, which does appear to be actual errors (it is recording its own errors, and it hurts performance), and I wanted to make sure that our 2960 switches are fine. The "Sender" switch's interface seems to always drop a tons of packets going to the "Receiver" switch. Second-2960X # show interfaces gigabitEthernet 1/0/48 counters errors Port Align-Err FCS-Err Xmit-Err Rcv-Err UnderSize OutDiscards So it's not an issue of saturating a line.įirst-Desktop (Sender) -> First-2960X -> Second-2960X -> Second-Desktop (Receiver)įirst-2960X # show interfaces GigabitEthernet 1/0/48 counters errors It doesn't matter if the transfer rate is is ~974 Mbps or ~480 Mbps or just ~279 Mbps. Let's say a user is copying a file from their computer to another computer. We mostly have 2960X switches (and one 3850). So I logged into some switches to check things out. I started setting up some new monitoring stuff at work, and when I enabled SNMP monitoring of my Cisco switches, I suddenly saw a bunch of "red" show up all over in InterMapper.
0 kommentar(er)
