Hello BIND Community,
I am writing to report a significant performance drop observed after upgrading
from BIND 9.18.30 to BIND 9.20.8 .
We are running BIND in a batch data processing environment where large volumes
of dynamic DNS updates are pushed periodically.
Under 9.18.30, our system consistently achieved a throughput of approximately
16,000 TPS (transactions per second).
However, after moving to 9.20.8, the throughput has dropped to around 5,000
TPS, despite using an identical configuration and environment.
Key Environment and Configuration Details:
* BIND Version: 9.20.8
* linked to OpenSSL version: OpenSSL 1.1.1k FIPS 25 Mar 2021
* compiled with libuv version: 1.48.0
* linked to libuv version: 1.48.0
* compiled with liburcu version: 0.15.0
* compiled with jemalloc version: 5.3.0
* compiled with libxml2 version: 2.9.7
* linked to libxml2 version: 20907
* compiled with zlib version: 1.2.11
* linked to zlib version: 1.2.11
* Platform: RHEL 8.5, GCC 8.5.0
named.configuration
options {
…
zone-statistics no;
notify yes;
max-cache-size 14297m;
max-journal-size 1048576;
…
};
ulimit -a
core file size (blocks, -c) 0
data seg size (kbytes, -d) unlimited
scheduling priority (-e) 0
file size (blocks, -f) unlimited
pending signals (-i) 254734
max locked memory (kbytes, -l) 64
max memory size (kbytes, -m) unlimited
open files (-n) 65536
pipe size (512 bytes, -p) 8
POSIX message queues (bytes, -q) 819200
real-time priority (-r) 0
stack size (kbytes, -s) 8192
cpu time (seconds, -t) unlimited
max user processes (-u) 65536
virtual memory (kbytes, -v) unlimited
file locks (-x) unlimited
cat /proc/mounts | grep nfs
sunrpc /var/lib/nfs/rpc_pipefs rpc_pipefs rw,relatime 0 0
Use case :
Trying to load 120M identifier of length 15 digit which will load in zone using
nsupdate in a batch of 1000.
A file created having 120M identifier and sent to nsupdate in a batch of 1000
to add identifier in bind.
Before Upgrade time taken to load data 02 hr 02min 02 second
After Upgrade time taken to load data 05 hr 5 min 27 second
Questions for the Community:
Has anyone else observed a similar TPS drop in 9.20.8 under update loads?
Are there internal configuration or behavioural changes in 9.20.8 affecting
update performance?
Are any workarounds or tunings recommended for high-performance batch DNS
update use cases in newer versions?
Is there any known regression or tuning guidance from ISC related to this?
Any input, confirmation, or suggestions would be highly appreciated. We are
open to providing additional logs or traces if needed.
Regards,
Sahil
--
Visit https://lists.isc.org/mailman/listinfo/bind-users to unsubscribe from
this list
ISC funds the development of this software with paid support subscriptions.
Contact us at https://www.isc.org/contact/ for more information.
bind-users mailing list
bind-users@lists.isc.org
https://lists.isc.org/mailman/listinfo/bind-users
