21 Říj
2019
21 Říj
'19
11:25
Hi Bastien,
Probably you have to increase
https://www.knot-dns.cz/docs/2.9/singlehtml/index.html#tcp-io-timeout
Daniel
On 10/21/19 11:20 AM, Bastien Durel wrote:
Hello,
I have a problem with a knot slave instance, which does not refresh
anymore.
When doing a retransfer, I have a "no usable master" error, with this
error log:
oct. 21 11:02:16 corrin knotd[32453]: info: [geekwu.org.] control, received command
'zone-retransfer'
oct. 21 11:02:16 corrin knotd[32453]: 2019-10-21T11:02:16 info: [geekwu.org.] control,
received command 'zone-retransfer'
oct. 21 11:02:17 corrin knotd[32453]: info: [geekwu.org.] AXFR, incoming, remote
10.42.42.6@53, starting
oct. 21 11:02:17 corrin knotd[32453]: 2019-10-21T11:02:17 info: [geekwu.org.] AXFR,
incoming, remote 10.42.42.6@53, starting
oct. 21 11:02:18 corrin knotd[32453]: debug: [geekwu.org.] refresh, remote master,
address 10.42.42.6@53, failed (connection reset)
oct. 21 11:02:18 corrin knotd[32453]: 2019-10-21T11:02:18 debug: [geekwu.org.] refresh,
remote master, address 10.42.42.6@53, failed (connection reset)
oct. 21 11:02:18 corrin knotd[32453]: 2019-10-21T11:02:18 warning: [geekwu.org.] refresh,
remote master not usable
oct. 21 11:02:18 corrin knotd[32453]: 2019-10-21T11:02:18 error: [geekwu.org.] refresh,
failed (no usable master)
oct. 21 11:02:18 corrin knotd[32453]: warning: [geekwu.org.] refresh, remote master not
usable
oct. 21 11:02:18 corrin knotd[32453]: error: [geekwu.org.] refresh, failed (no usable
master)
On master, I have this error log:
oct. 21 11:02:16 arrakeen knotd[20289]: info: [geekwu.org.] AXFR, outgoing, remote
10.120.42.42@33435, started, serial 2019100808
oct. 21 11:02:16 arrakeen knotd[20289]: 2019-10-21T11:02:16 info: [geekwu.org.] AXFR,
outgoing, remote 10.120.42.42@33435, started, serial 2019100808
oct. 21 11:02:16 arrakeen knotd[20289]: debug: TCP, send, address 10.120.42.42@33435
(connection timeout)
oct. 21 11:02:16 arrakeen knotd[20289]: 2019-10-21T11:02:16 debug: TCP, send, address
10.120.42.42@33435 (connection timeout)
tcpdump shows a TCP connection from 10.120.42.42 to 10.42.42.6, which
is closed with a TCP FIN by 10.42.42.6 after 1.2s
Do you have hints about that ?
Thanks,