IP fragmentation fails and fragmented packets get dropped

Solution Verified - Updated -

Issue

  • We have a three node Oracle RAC and this morning we had problem with one node. When we reviewed the historic netstat -s ouput we noticed that the following counter spiked around that time.
 54672220 packet reassembles failed
  • After that event the module could not be pinged with a packet size greater than the MTU.
# ping <IP Addr> -s 1600

Environment

  • Red Hat Enterprise Linux 6.6 (kernel-2.6.32-504.el6) and later
  • Red Hat Enterprise Linux 6.7.z (kernel-2.6.32-573.7.1.el6) and earlier
  • Oracle RAC or other application which commonly sends IP fragments
  • IP fragmentation
  • Systems with many CPU cores
    • In one instance, the system had 160 CPU cores
  • Sytemss with smaller amounts of CPUs are not affected.
    • For example, we could not reproduce this on a machine with 12 CPU cores

Subscriber exclusive content

A Red Hat subscription provides unlimited access to our knowledgebase of over 48,000 articles and solutions.

Current Customers and Partners

Log in for full access

Log In