No subject


Tue Apr 2 00:56:56 UTC 2013


period of about one month the cache gradually grew larger and larger,
until it eventually started eating into swap and affected the response
time of lookups.

Tim, did you get your E250's working properly in the end? (with regards
to network errors).

Also, just noticed Brad's posting as I was writing this.
>Why not install local caching servers on each machine, so 
>that as you add more clients to do DNS traffic, you also add more 
>horsepower to resolve that DNS traffic, and your nameservices scale 
>linearly with the machines?

Sorry Brad, but can you explain this configuration a little, I found it
hard to work out what you meant, sounds good though ;)


Chris.
--
further details to my previous post:
http://groups.google.com/groups?q=BIND+9.1.2+memory+usage&hl=en&lr=&safe=off&rnum=1&ic=1&selm=9dh8et%242or6%241%40FreeBSD.csie.NCTU.edu.tw
(BIND 9.1.1, FreeBSD 4.2, with 256MB RAM, 512MB swap)

Status of named:
USER    PID %CPU %MEM   VSZ  RSS  TT  STAT STARTED      TIME COMMAND
6660   6078 49.9 89.1 276940 230704  ??  RsJ  12Apr01 6604:27.50
/usr/local/sbin/named -u named

rndc: stats command successful
+++ Statistics Dump +++ (989313994)
success 489315967
referral 92547
nxrrset 5084719
nxdomain 39244607
recursion 30722105
failure 12177661
--- Statistics Dump --- (989313994)

su-2.04# top
last pid: 29782;  load averages:  0.53,  0.54,  0.46 up 26+22:38:57 
10:27:43
17 processes:  2 running, 15 sleeping
CPU states: 45.7% user,  0.0% nice,  9.8% system,  1.5% interrupt, 43.0%
idle
Mem: 219M Active, 8444K Inact, 15M Wired, 6584K Cache, 35M Buf, 660K
Free
Swap: 516M Total, 132M Used, 384M Free, 25% Inuse, 440K In, 128K Out

  PID USERNAME PRI NICE  SIZE    RES STATE    TIME   WCPU    CPU COMMAND
 6078   6660    45   0   270M   226M RUN    110.1H 53.32% 53.32% named


More information about the bind-users mailing list