Sunghwan Kim(IBI)
2018-09-14 00:09:31 UTC
Hi BIND expert,
I could not have sent the followings thru https://www.isc.org/bind-
subscription-contact/
due to error on the site.
--------------
I am a S/W engineer who is working on BIND, especially named in Seoul/Korea.
I've got reports from a customer regarding stucked "named" process which had
not been performed any request from clients for 5 secs during "rndc
reconfig"
even if it is used to be finished in 700ms
24-Aug-2018 17:36:39.073 general: info: received control channel command
'reconfig'
¡Š..
24-Aug-2018 17:36:44.100 general: info: any newly configured zones are now
loaded
My customer's DNS server has been installing BIND-9.9.11-P1.
I would like to figure out why named was stucked for even 5 secs on rndc
reconfig.
I've figured out I/O event values(majflt/s) of SAR information on the
server is quite high
which is 58.34 even if it usually is 0.18 ~ 0.32.
The server information is as following;
1. OS : CentOS 7.3
2. CPU : Intel Xeon3.5Ghz 64bits(6 CPUs, 2 cores per CPU)
3. Mem. : 8G
Would you please give me any information about it ?
I know a lot of fixes on ¡°rndc reconfig¡± for latter version of 9.9.11-P1
Please take a look at the following logs from bind for your information;
=== general log =====
24-Aug-2018 17:36:39.073 general: debug 1: received control channel command
'null'
24-Aug-2018 17:36:39.073 general: info: received control channel command
'reconfig'
24-Aug-2018 17:36:39.073 general: info: loading configuration from
'/etc/named.conf'
24-Aug-2018 17:36:39.159 general: info: unable to open
'conf/named.iscdlv.key' using built-in keys
24-Aug-2018 17:36:39.168 general: info: using default UDP/IPv4 port range:
[9000, 61000]
24-Aug-2018 17:36:39.169 general: info: using default UDP/IPv6 port range:
[9000, 61000]
24-Aug-2018 17:36:39.190 general: info: sizing zone task pool based on 4704
zones
24-Aug-2018 17:36:39.293 general: debug 1: zone_settimer: zone xn--
pi5bm5e/IN: enter
¡Š.(removed)¡Š..
24-Aug-2018 17:36:41.809 general: debug 1: zone_settimer: zone xn--o78b/IN:
enter
24-Aug-2018 17:36:41.816 general: info: dns64 reverse zone: 0.0.0.0.0.0.0.0.
0.0.0.0.0.0.0.0.b.9.f.f.4.6.0.0.ip6.arpa.
¡Š.(removed)¡Š..
24-Aug-2018 17:36:43.927 general: debug 1: now using logging configuration
from config file
24-Aug-2018 17:36:43.935 general: info: zone
0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.b.9.f.f.4.6.0.0.ip6.arpa/IN: (master)
removed
24-Aug-2018 17:36:43.938 general: debug 1: load_configuration: success
24-Aug-2018 17:36:43.938 general: info: reloading configuration succeeded
¡Š¡Š¡Š¡Š
It would be appreciated if you share any hints, information.
Regards,
Sunghwan.
------------------------------
(ÁÖ)ŸÆÀ̺ñŸÆÀÌ(www.ibi.net)
DNS»çŸ÷ºÎ/º»ºÎÀå
02-2165-7234/010-3558-3736
[03909]Œ¿ï ž¶Æ÷±ž žÅºÀ»ê·Î 37(»óŸÏµ¿, DMC»êÇÐÇù·ÂŒŸÅÍ1304È£)
I could not have sent the followings thru https://www.isc.org/bind-
subscription-contact/
due to error on the site.
--------------
I am a S/W engineer who is working on BIND, especially named in Seoul/Korea.
I've got reports from a customer regarding stucked "named" process which had
not been performed any request from clients for 5 secs during "rndc
reconfig"
even if it is used to be finished in 700ms
24-Aug-2018 17:36:39.073 general: info: received control channel command
'reconfig'
¡Š..
24-Aug-2018 17:36:44.100 general: info: any newly configured zones are now
loaded
My customer's DNS server has been installing BIND-9.9.11-P1.
I would like to figure out why named was stucked for even 5 secs on rndc
reconfig.
I've figured out I/O event values(majflt/s) of SAR information on the
server is quite high
which is 58.34 even if it usually is 0.18 ~ 0.32.
The server information is as following;
1. OS : CentOS 7.3
2. CPU : Intel Xeon3.5Ghz 64bits(6 CPUs, 2 cores per CPU)
3. Mem. : 8G
Would you please give me any information about it ?
I know a lot of fixes on ¡°rndc reconfig¡± for latter version of 9.9.11-P1
Please take a look at the following logs from bind for your information;
=== general log =====
24-Aug-2018 17:36:39.073 general: debug 1: received control channel command
'null'
24-Aug-2018 17:36:39.073 general: info: received control channel command
'reconfig'
24-Aug-2018 17:36:39.073 general: info: loading configuration from
'/etc/named.conf'
24-Aug-2018 17:36:39.159 general: info: unable to open
'conf/named.iscdlv.key' using built-in keys
24-Aug-2018 17:36:39.168 general: info: using default UDP/IPv4 port range:
[9000, 61000]
24-Aug-2018 17:36:39.169 general: info: using default UDP/IPv6 port range:
[9000, 61000]
24-Aug-2018 17:36:39.190 general: info: sizing zone task pool based on 4704
zones
24-Aug-2018 17:36:39.293 general: debug 1: zone_settimer: zone xn--
pi5bm5e/IN: enter
¡Š.(removed)¡Š..
24-Aug-2018 17:36:41.809 general: debug 1: zone_settimer: zone xn--o78b/IN:
enter
24-Aug-2018 17:36:41.816 general: info: dns64 reverse zone: 0.0.0.0.0.0.0.0.
0.0.0.0.0.0.0.0.b.9.f.f.4.6.0.0.ip6.arpa.
¡Š.(removed)¡Š..
24-Aug-2018 17:36:43.927 general: debug 1: now using logging configuration
from config file
24-Aug-2018 17:36:43.935 general: info: zone
0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.b.9.f.f.4.6.0.0.ip6.arpa/IN: (master)
removed
24-Aug-2018 17:36:43.938 general: debug 1: load_configuration: success
24-Aug-2018 17:36:43.938 general: info: reloading configuration succeeded
¡Š¡Š¡Š¡Š
It would be appreciated if you share any hints, information.
Regards,
Sunghwan.
------------------------------
(ÁÖ)ŸÆÀ̺ñŸÆÀÌ(www.ibi.net)
DNS»çŸ÷ºÎ/º»ºÎÀå
02-2165-7234/010-3558-3736
[03909]Œ¿ï ž¶Æ÷±ž žÅºÀ»ê·Î 37(»óŸÏµ¿, DMC»êÇÐÇù·ÂŒŸÅÍ1304È£)