[dns-operations] Truncated, retrying in TCP mode

[Subramanian, Karthikeyan]

Hi All,

Need suggestion.

we are seeing delay in TCP query for 4sec in some set of caching servers (Its acting as a forwarder).
The query is having bunch of NS (61 Name server) so its truncating to TCP.  Only the delay is happening on particular set of forwarders.

OS
Linux 2.6.32-642.el6.x86_64

I got this option from one site to set the “net.ipv4.tcp_limit_output_bytes” to 128K (https://lwn.net/Articles/507065/)
The default value of this limit is 128KB; it could be set lower on systems where latency is the primary concern.
/proc/sys/net/ipv4/tcp_limit_output_bytes
- net.ipv4.tcp_limit_output_bytes = 262144
+ net.ipv4.tcp_limit_output_bytes = 131072

Need advice on this.

Thanks
Karthikeyan.S


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.dns-oarc.net/pipermail/dns-operations/attachments/20160726/3ee79fa3/attachment.html>