Hi,
That is quite an old version now. Might be worth updating to a newer version as a side note :-)
For 5.4.x Both probing and threshold defaults are 1 - So, if 1 packet is missed/dropped it goes inactive and then back up again when first success is seen.
Depending on how you manage your network, I would personally add/change the params: (disclaimer: YMMV though)
modparam("dispatcher", "ds_probing_threshold", 3)
modparam("dispatcher", "ds_ping_interval", 3)
modparam("dispatcher", "ds_inactive_threshold", 1)
The above requires 3 OPTIONS to be dropped before marking inactive with a packet sent every 3 seconds. (9 seconds before Flag: IP) It might give you a buffer for odd networking issues.
Back to your original question though, if the destination dropped said OPTIONS packet and didn't come back online until the restart of the service that means that due to param below - it means it was constantly trying it.
modparam("dispatcher", "ds_probing_mode", 1)
This could be a weird network/routing/firewall/congestion type issue, but it is impossible for us to know for sure.
The above params should help but you can tune it further if your boxes are busy, like using:
ds_timer_mode &
ds_ping_latency_stats both set to 1
Have a read of the docs and have a play around with the params, they are really helpful.
Thanks,
John.
Hi,
Thanks for the reply,
The dispatcher params we are using are as below:
#--- Dispatcher params ----
modparam("dispatcher", "db_url", DBURL_LOCAL)
modparam("dispatcher", "flags", 2)
modparam("dispatcher", "xavp_dst", "ds_dst")
modparam("dispatcher", "ds_probing_mode", 1)
modparam("dispatcher", "ds_ping_interval", 30)
modparam("dispatcher", "ds_ping_from", "sip:proxy@VAR_LOCAL_IP")
The version of the kamailio is 5.4.7.
Above are the details which was asked, Please let me know if any thing is needed.
Hi,
Check the options you have set for the dispatcher params and then if you have any logic to mark dispatcher destination as IP on any timeout, or error response would be my first checks.
Specifically - "ds_probing_threshold" and the "ds_inactive_threshold" params (first takes it out after x failed, second adds it back in after x successful.) We would need to see the kamailio version, dispatcher params and the dispatcher logic to help further though.
Thanks,
John.
HI,
Scenario:
We possess 2 proxies and 2 registrar servers equipped with dmq functionality. Unexpectedly, one of the Proxy IP addresses was marked as inactive (FLAG: IP) in the dispatcher list, causing the proxy to cease responding. Simultaneously, Kamailio logs stopped recording any activity.
Upon restarting the Kamailio service, the problem was resolved, and the proxy resumed normal operation, with the Flag changing back to AP (ACTIVE).
The sudden occurrence of this issue is not clear, and due to the cessation of logs during the incident, we lack information for the investigation.
Any suggestions on how to address this issue would be appreciated.
And is this scenario happened to any one in your experience?
__________________________________________________________
Kamailio - Users Mailing List - Non Commercial Discussions
To unsubscribe send an email to sr-users-leave@lists.kamailio.org
Important: keep the mailing list in the recipients, do not reply only to the sender!
Edit mailing list options or unsubscribe: