what causes flushes?

This rural ISP is having problems described below; any help would be appreciated.

Configuration: cisco 7206VXR (NPE300) processor with 229376K/65536K bytes of memory IOS (tm) 7200 Software (C7200-K4P-M), Version 12.0(32)S6, RELEASE SOFTWARE (fc3)

she-gw#show process cpu | incl utilization CPU utilization for five seconds: 68%/46%; one minute: 67%; five minutes: 66%

Issue: poor Miranda

formatting link

results from customer's viewpoint, including lower downstream rates than upstream rates; typically 3% of packets dropped when unix host pings unix host on another router interface.

1st question: should we be concerned about the number of flushes below?

kind regards/ldv

Larry Vaden Internet Texoma, Inc.

she-gw#show interface Serial3/0 | incl counter Last clearing of "show interface" counters 01:00:09 she-gw#show interface | incl Input queue: Input queue: 1/4096/0/51372 (size/max/drops/flushes); Total output drops: 0

---FE0/0, 18 subnets (secondaries) Input queue: 0/75/0/0 (size/max/drops/flushes); Total output drops:

0 Input queue: 0/75/0/0 (size/max/drops/flushes); Total output drops: 0 Input queue: 0/75/0/0 (size/max/drops/flushes); Total output drops: 0 Input queue: 0/75/0/22 (size/max/drops/flushes); Total output drops: 0 Input queue: 0/4096/0/0 (size/max/drops/flushes); Total output drops: 0 Input queue: 0/75/0/0 (size/max/drops/flushes); Total output drops: 0 Input queue: 0/75/0/29 (size/max/drops/flushes); Total output drops: 85 Input queue: 0/75/0/0 (size/max/drops/flushes); Total output drops: 0 Input queue: 0/75/0/2272 (size/max/drops/flushes); Total output drops: 0 Input queue: 24/75/49/107221 (size/max/drops/flushes); Total output drops: 0

---Serial3/0, upstream DS3 Input queue: 0/75/0/501 (size/max/drops/flushes); Total output drops: 0 Input queue: 0/75/0/1778 (size/max/drops/flushes); Total output drops: 0 Input queue: 0/75/0/0 (size/max/drops/flushes); Total output drops:

0 she-gw

ha scritto nel messaggio news: snipped-for-privacy@z28g2000prd.googlegroups.com...

Hi,

Flushes generally occur when there is an excessive packet processing for your Route Processor. Some "user" packets are to be discarded to let priority packets be processed (e.g. control packets ( keepalives, routing protocols traffic, other management traffic). Sometimes a consistent number of flushed data is associated with some kind of DoS attack.

Note that this protection come in place only with Process Switching; with CEF enabled only a small percentage of traffic should be punted to the RP.

a) Assess why so many packets are process switched b) Try and increase your input queue c) Upgrade your route processor

Regards, Gabriele