Hi, I have raised the issues I experienced below with Mikrotik support but thought I'd put it here too for any comment . . .
We are performing some testing on the mikrotik os7 with a view to implementing this on our customer sites going forward.
We currently use os6.
We have tested using combinations of the following Mikrotik hardware
Routerboard RB3011 UiAS-RM
Routerboard 1100AHx4
CCR2116 12G 4S+
After several days of testing, we have concluded that the PIM SM implementation in os7 is extremely buggy!
We have tested with versions upto 7.10.2 including the beta version 7.11rc1
I created a simple test to reproduce some of the problems we experienced. We normally use GRE tunnels, OSPF and BSR/RP candidates - but for simplicity in this test, we setup static routes and static RP over ethernet.
SCENARIO 1
Site1
Ether1 - 10.1.1.1/24
Ether6 - 192.168.1.1/24
Route 192.168.2.0/24 via 10.1.1.2
PIM SM with default settings
Static RP 10.1.1.1
Site2
Ether1 - 10.1.1.2/24
Ether6 - 192.168.2.1/24
Route 192.168.1.0/24 via 10.1.1.1
PIM SM with default settings
Static RP 10.1.1.1
The two sites Ether1 ports are connected via a common switch
Step1
We start a multicast source on a system connected to Site2 (src: 192.168.2.254 grp: 233.75.215.2)
We start a multicast receiver on a system connected to Site1 (for grp 233.75.215.2)
We observe the multicast traffic being received at Site1
Step2
We unplug the network cable from the ether1 port at Site1
After a 5 second wait, we reconnect the cable
Observation
The multicast traffic does not resume.
We observe thousands and thousands of records being generated in the log at Site1 as shown in the example below.
In this condition, we are unable to access to any of the PIM settings in the GUI or terminal session on Site1
Interface {0 IP} created
Interface {0 IP} add static rp for 224.0.0.0/4 10.1.1.1
Interface {0 IP} interface { ether 1} created
Interface {0 IP} interface { ether 6} created
Interface {0 IP} uib {grp: 233.75.215.2 } created
Interface {0 IP} uib {src: 192.168.2.254 grp: 233.75.215.2 } created
Interface {0 IP} created
Interface {0 IP} add static rp for 224.0.0.0/4 10.1.1.1
Interface {0 IP} interface { ether 1} created
Interface {0 IP} interface { ether 6} created
Interface {0 IP} uib {grp: 233.75.215.2 } created
Interface {0 IP} uib {src: 192.168.2.254 grp: 233.75.215.2 } created
Interface {0 IP} created
Interface {0 IP} add static rp for 224.0.0.0/4 10.1.1.1
Interface {0 IP} interface { ether 1} created
Interface {0 IP} interface { ether 6} created
Interface {0 IP} uib {grp: 233.75.215.2 } created
Interface {0 IP} uib {src: 192.168.2.254 grp: 233.75.215.2 } created
. . . . this loop continues indefinitely (> 24 hours) until the source multistream is stopped.
Once the source multicast stream is stopped, we can then access the PIM settings on Site1.
SCENARIO 2
Same setup configuration as scenario 1
Step1
The two routers are connected with no multicast streams established
We observe in Site1/routing/PIM SM/Neighbor that we have an entry for the Site2 router 10.1.1.2 (as expected)
Step2
The ether1 port on the Site2 router is unplugged
After the PIM neighbor timer expires, we see the following entry in the log (as expected)
Interface {0 IP} Interface { ether1 } neighbor {10.1.1.2 } destroyed
Step3
The ether1 port on the Site2 router is reconnected
Step4
We start a multicast source on a system connected to Site2 (src: 192.168.2.254 grp: 233.75.215.2)
We start a multicast receiver on a system connected to Site1 (for grp 233.75.215.2)
We observe the multicast traffic being received at Site1
Step5
The ether1 port on the Site2 is unplugged
After the PIM timer expires, we DO NOT SEE the neighbor entry get removed.
Immediately after this point, as the PIM neighbor timeout hits zero, we are subsequently unable to access to any of the PIM settings in the GUI or a terminal session.
Looking forward to any ideas