OSPFv2 Graceful Restart - interoperability

[rchandrans]

Switch running FRR is connected to a Arista switch.
FRR is configured as GR restarting node and Arista configured as GR helper
After graceful restart is induced on FRR, grace lsa's are flooded and the helper node has received the grace-lsa.
OSPF session on the helper node is FULL even after FRR software is reloaded.
After FRR is up after sw reload, OSPF session is stuck in Exstart state until grace period expires.
Did a packet capture to see the control packet exchange.

1. OSPF DD packets sent from FRR is not acknowledged by Arista
2. OSPF DD packets are re-transmitted by FRR
3. Below log is seen in Arista.
   Jan 18 18:59:32 CBM-ARISTA-7280R Rib: OSPF: invalid DD packet from 32.32.32.1: Master flag mismatch (14)
   Jan 18 19:00:07 CBM-ARISTA-7280R Rib: last message repeated 7 times
   Jan 18 19:01:12 CBM-ARISTA-7280R Rib: last message repeated 13 times
   Jan 18 19:02:13 CBM-ARISTA-7280R Rib: last message repeated 12 times
   Jan 18 19:03:13 CBM-ARISTA-7280R Rib: last message repeated 12 times
4. checked MTU and it does not appear to be an MTU issue.
5. Issue is not observed if, restarting node and helper node are both FRR.
6. Checked frr code, master/slave for DD packet is established based on router id( higher router id ip wins mastership for dd packet).
   Looking forward for pointers.