Thanks to Sandeep we managed to get access to the test node in RDO infra to look into the issue.
I'll summarize our observations below.
* `iptables -t filter -L INPUT` takes almost 40 seconds in the node[1].
* The command gets stack for several seconds before it shows a line with 192.168.24.0/24.
* adding -n option allows the command to complete quickly[1].
* We tested removing 127.0.0.1 from resolve.conf, which resolved the slowness without -n
Because ansible_iptables module executes the command when it inserts a single rule, the slow command can cause huge delay when there are multiple rules managed.
At this moment we are concluding the issue is caused by the wrong DNS setting, which we'll look into further.
Thanks to Sandeep we managed to get access to the test node in RDO infra to look into the issue.
I'll summarize our observations below.
* `iptables -t filter -L INPUT` takes almost 40 seconds in the node[1].
* The command gets stack for several seconds before it shows a line with 192.168.24.0/24.
* adding -n option allows the command to complete quickly[1].
* We tested removing 127.0.0.1 from resolve.conf, which resolved the slowness without -n
[1] https:/ /paste. opendev. org/show/ beGwwFUdaaaO0R2 ccOdy/
[2]
nameserver 127.0.0.1
nameserver 1.1.1.1
Because ansible_iptables module executes the command when it inserts a single rule, the slow command can cause huge delay when there are multiple rules managed.
At this moment we are concluding the issue is caused by the wrong DNS setting, which we'll look into further.