Hi,I am trying to create multiple containers in one host and each connected to a PPTP server,disconnect and reconnect, repeating
When I created 10 containers , these pptp clients worked fine, but after I created 50 containers, many pptp clients could not connect to the server. Here is the log:
Using interface ppp0
Connect: ppp0 <--> /dev/pts/4
LCP: timeout sending Config-Requests
Connection terminated.
Modem hangup
It looks like a GRE problem, but when I use tcpdump to monitor traffic, tcpdump does not detect any packets except DNS requests, like the following log:
# start with a dns request
14:05:08.364915 IP a70ba599f847.41336 > public1.alidns.com.53: 4216+ A? myvpnserver.com. (39)
14:05:08.379093 IP public1.alidns.com.53 > a70ba599f847.41336: 4216 1/0/0 A 182.x.x.x (55)
14:05:13.369424 ARP, Request who-has 172.17.0.1 tell a70ba599f847, length 28
14:05:13.369460 ARP, Reply 172.17.0.1 is-at 02:42:a8:46:83:5d (oui Unknown), length 28
# no more packet untill timeout
#sometimes got a ARP from host
14:03:05.209453 ARP, Request who-has a70ba599f847 tell 172.17.0.1, length 28
14:03:05.209462 ARP, Reply a70ba599f847 is-at 02:42:ac:11:00:02 (oui Unknown), length 28
Some PPTP clients have a probability of connecting properly, but most of them cannot be connected.
My container uses a bridge network. The total number of containers is 50.
When the total number of containers is 10, everything works fine.
BTW, when the PPTP client cannot connect, Ping PPTP server is working find on container.
Does anyone know the relevant information?
Host: ubuntu16.04
Docker version 18.09.7, build 2d0083d
Thanks
lth