Explanation of "mroute: pending queue full. dropping entries" event in messages file

book

Article ID: 167980

calendar_today

Updated On:

Products

XOS

Issue/Introduction

This article explain the reason for the event "mroute: pending queue full. dropping entries" in /var/log/messages fileThe following kernel event may appear in /var/log/messages file when using PIM to route multicast traffic:

Sep 15 19:23:36 fw_1 kernel: mroute: pending queue full, dropping entries.

Some of the mroute entries show "Register" as the outgoing interface in NSM. For example:

NSM# show ip mroute 

IP Multicast Routing Table 
Flags: I - Immediate Stat, T - Timed Stat, F - Forwarder installed 
Timers: Uptime/Stat Expiry 
Interface State: Interface (TTL) 

(192.168.1.1, 226.10.10.10), uptime 00:05:45, stat expires 00:02:03 
Owner PIM-SM, Flags: TF 
Incoming interface: inside 
Outgoing interface list: 
Register (1) 

Cause

Problem:

When a streamer start sending its' video channel, the first hop router registers a new multicast group with the Rendezvous Point (RP).

Until the router receives a confirmation from the RP (PIM register-stop signal), the mroute entry is in the Register state and clients can not join the multicast group.



Resolution

There can be various reasons for a channel not to get registered with the RP.
  • Check if IP TTL setting on the streamer is sufficient to pass every L3 hop until it reaches RP.
  • Check if the relevant VAP group has the correct unicast routing towards the RP.
  • Check if pim-register messages leave the VAP and if pim register-stop messages from the RP reach the VAP. 
There are different types of network services that send multicast traffic not supposed to be joined externally.

For example, Microsoft Windows Server clusters use multicast messages and  some of them are using TTL equal to 1. If a VAP group running PIM is a part of that network, it will receive the traffic and try to register it with the RP.

If multicast support isn't required on the circuit connected to such network, you can disable "ip pim sparse-mode" in the PIM circuit configuration. Alternatively, you can create a flow-rule with action drop for that specific multicast traffic so it doesn't reach the VAP group.

Workaround

N/A