EIGRP SIA -unknown drawbacks of the DUAL FSM ??

[shiladitya012]

Hi all,

while reading EIGRP  found that the most problematic thing about this protocol is SIA . I know the FSM and all the Inputs Events related to it as well.

But why EIGRP ends up in SIA for the outstanding Q/R ?

Is it related to unknown any drawbacks of the DUAL FSM ?

is there any possibility where State transition mechanism of DUAL is lacking a inbuilt recovery from SIA ?

[Reza Sharifi]

Hi,

See this link to Peter's answer on EIGRP and SIA.

https://supportforums.cisco.com/discussion/12034081/what-does-eigrp-stuck-inactive-state-mean

HTH

[shiladitya012]

Hi,

the link is very helpful . it describes how and when SIA happens.

But when EIGRP process is running for a lost route it places the route in ACTIVE mode.As Peter said that " By default, a SIA state is declared if the diffusing computation could not be completed in 3 minutes (180 seconds)." so the SIA route remains the same way in the topology table for eternity . in the mean time even if the neighbor sends a REPLY to the last Query or sends another Update for the lost route the Active router won't install the information in Topology table. To use the information we need to manually clear SIA topology information from Topology table and then re-sync the neighborship.

I have seen several cases where the Active router is receiving valid information from the neighbor but unable to use it as the route is in SIA state . 

So is there any possibility where State transition mechanism of DUAL is lacking a inbuilt recovery from SIA ? why we have to manually clear the topology table and re-sync the neighborship ?