if input / output drops

Recently, an SNMP monitoring program has been displaying errors for my backbone E1. When the link is saturated, downstream traffic gets input / output packet drops.

Is it a smart thing to increase the hold-queue? or disable CEF?

Platform: Cisco7206VXR

Serial2/0 is up, line protocol is up Hardware is 4ME1-BAL Description: E1#1 8:1/2 i8 Internet address is 10.10.10.34/30 MTU 1500 bytes, BW 1984 Kbit, DLY 20000 usec, reliability 255/255, txload 92/255, rxload 42/255 Encapsulation HDLC, crc 16, loopback not set Keepalive set (10 sec) Restart-Delay is 0 secs Last input 00:00:00, output 00:00:00, output hang never Last clearing of "show interface" counters 2w1d Input queue: 0/75/0/15 (size/max/drops/flushes); Total output drops:

12266 Queueing strategy: weighted fair Output queue: 1/1000/64/12266 (size/max total/threshold/drops) Conversations 1/247/256 (active/max active/max total) Reserved Conversations 0/0 (allocated/max allocated) Available Bandwidth 1488 kilobits/sec 5 minute input rate 330000 bits/sec, 169 packets/sec 5 minute output rate 723000 bits/sec, 162 packets/sec 175347945 packets input, 1677577210 bytes, 0 no buffer Received 151254 broadcasts, 0 runts, 0 giants, 0 throttles 862300 input errors, 861883 CRC, 0 frame, 7 overrun, 0 ignored, 410 abort 143679042 packets output, 2524635557 bytes, 0 underruns 0 output errors, 0 collisions, 2 interface resets 0 output buffer failures, 0 output buffers swapped out 2 carrier transitions 0 alarm indications, 0 remote alarms, 0 rx LOF, 0 rx LOS DCD up, BER inactive, NELR inactive, FELR inactive

Serial2/0 is up, line protocol is up Internet address is 10.10.10.34/30 Broadcast address is 255.255.255.255 Address determined by non-volatile memory MTU is 1500 bytes Helper address is not set Directed broadcast forwarding is disabled Multicast reserved groups joined: 224.0.0.10 Outgoing access list is not set Inbound access list is not set Proxy ARP is enabled Security level is default Split horizon is enabled ICMP redirects are always sent ICMP unreachables are always sent ICMP mask replies are never sent IP fast switching is enabled IP fast switching on the same interface is enabled IP Flow switching is disabled IP CEF switching is enabled IP Feature Fast switching turbo vector IP Feature CEF switching turbo vector IP multicast fast switching is enabled IP multicast distributed fast switching is disabled IP route-cache flags are Fast, CEF Router Discovery is disabled IP output packet accounting is disabled IP access violation accounting is disabled TCP/IP header compression is disabled RTP/IP header compression is disabled Probe proxy name replies are disabled Policy routing is disabled Network address translation is disabled WCCP Redirect outbound is disabled WCCP Redirect inbound is disabled WCCP Redirect exclude is disabled BGP Policy Mapping is disabled IP multicast multilayer switching is disabled

Reply to
hack.bac
Loading thread data ...

If the link is saturated, there is nothing you can do to prevent that except increase the amount of bandwidth between A and B. The reason you are getting output drops is because you are trying to send more traffic than the bandwidth you have can send. Input drops are a different story, but the number of input drops you have is not an area of concern.

CEF is a method to populate the hardware routing cache, it has nothing to do with input/output drops, except that disabling it will result in more drops because it will significantly slow routing performance.

Scott

Reply to
Thrill5

Try to disable WFQ - use FIFO on high speed interfaces. increasing hold queue should be your last resort.

Reply to
jvolter

A couple of things here:

More concerning than the drops are the CRC errors. Maybe though they are historic?

Also, in general increasing buffer sizes is a bad thing. I know it sounds like a good thing but it is usually not and will never I don't think be a good thing of the line is saturated.

In the saturated case what you get by adding more output buffer is that you change from

- A saturated line

to

- A saturated line with more delay.

I am now very much against large buffers.

Here is an little analogy.

Imagine that the local supermarkets were required by law to do queuing in the same way as routers do it. The queue is hidden from the customer in a big lightless box such that customers (packets) cannot at any instant determine the size of the queue. Also if the queue is full customers are turned away by security at the queue entrance and ejected from the supermarket.

If the supermarket is busy enough you will be able to find information on the instantaneous and maximum queue size by measuring the time spent queueing at each visit.

I don;t know about you but I would be tempted to find out the relative 'hidden' queue sizes and go to the one with the smallest hidden queue.

If then the stores were required to announce the Maximum queue size (number of output packet buffers) I woudl be drawn to the one with the smaller number.

Good luck.

Reply to
Bod43

Cabling-Design.com Forums website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.