I've been experiencing an odd problem with a pair of 2851 routers linked via ISDN BRI. The dialer interfaces come up as expected and they bring up additional ISDN channels as the load increases. The problem occurs when either end drops an isdn channel due to low utilisation. I've pasted below the BRI configuration for the interfaces at each end:
interface BRI1/0 no ip address encapsulation ppp load-interval 30 dialer pool-member 10 dialer pool-member 20 isdn switch-type basic-net3 isdn point-to-point-setup *** this line seems to be added automatically *** ppp authentication chap ppp multilink
The same configuration is replicated on the remaining 3 BRI interfaces on both routers. The dialer interface on the head office router is below:
interface Dialer10 ip address 10.0.0.1 255.255.255.0 encapsulation ppp dialer pool 10 dialer remote-name remote dialer order round-robin dialer idle-timeout 7200 dialer enable-timeout 3 dialer wait-for-carrier-time 10 dialer string 900441923123456 dialer caller 900441923123456 callback dialer load-threshold 10 either dialer-group 5 ppp authentication chap ppp chap hostname central ppp multilink ppp multilink interleave
The remote end dialer config is below:
interface Dialer10 ip address 10.0.0.2 255.255.255.0 encapsulation ppp dialer pool 10 dialer remote-name central dialer order round-robin dialer idle-timeout 7200 dialer string 900441923654321 dialer load-threshold 10 either dialer-group 5 ppp authentication chap ppp chap hostname remote ppp multilink ppp multilink interleave
The strange thing is that when the problem occurs, each router can ping the router at the other end (both dialer and ethernet interfaces); however hosts at each site can no longer ping each other.
If I shut & no shut the dialer interface at either end this resolves the issue. I've put a workaround in place for the moment by specifying 'ppp multilink links minimum 8' with the same set as the maximum. This avoids loss of connectivity; however it's clearly costly...
With 'debug dialer events' enabled, I can see that the routers are dropping connections simply due to low threshold, which seems right.
For info, I'm running the following image at both ends: c2800nm- advsecurityk9-mz.124-10a.bin
Cheers in advance for any help, Chris