Hello again, This is my main problem : I can't use "network" mode because we already use user namespaces on these machines. And it's not compatible. Some posts here and there suggested BasicAddressResolver could be solution, but no example was provided and I am not really sure about what is offers and what it doesn't. So I gave a try, but it seems it is not a workaround. If there is no way to cluster it with NAT involved, I will have to find another database. Regards
Le 07-Dec-2022 13:38:50 +0100, [email protected] a crit: Thank you for the valuable information. "with forced IP 192.168.1.99" -- I think this is the reason. Ignite node does not know that its real IP is 10.17.10.55. If it is possible to use host docker network I believe it should help. https://docs.docker.com/network/host/ So, try to run both docker containers with --network host. -- Best regards, Aleksandr On 2022/12/07 10:35:13 [email protected] wrote: > Hi, 10.17.10.55 and 10.17.10.56 are the addresses of the VMs hosting the > containers. > My Ignite containers are based on the official image. So when I am logged > (with "docker exec") into the one hosted on 10.17.10.55, with forced IP > 192.168.1.99 (as it is not possible to force an address withing the "default" > range) : bash-5.1# hostname > aac1698dd409 bash-5.1# ip a > 1: lo: mtu 65536 qdisc noqueue state UNKNOWN qlen 1000 > link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 > inet 127.0.0.1/8 scope host lo > valid_lft forever preferred_lft forever > 11982: eth0@if11983: mtu 1500 qdisc noqueue state UP > link/ether 02:42:c0:a8:01:63 brd ff:ff:ff:ff:ff:ff > inet 192.168.1.99/16 brd 192.168.255.255 scope global eth0 > valid_lft forever preferred_lft forever > bash-5.1# ping 10.17.10.56 > PING 10.17.10.56 (10.17.10.56): 56 data bytes > 64 bytes from 10.17.10.56: seq=0 ttl=63 time=0.413 ms > 64 bytes from 10.17.10.56: seq=1 ttl=63 time=0.352 ms > bash-5.1# echo "10800 11211 47100 47500" | xargs -n 1 nc -vz 10.17.10.56 > 10.17.10.56 (10.17.10.56:10800) open > 10.17.10.56 (10.17.10.56:11211) open > 10.17.10.56 (10.17.10.56:47100) open > 10.17.10.56 (10.17.10.56:47500) open It seems ports exposed by the container > hosted by the other VM can be reached. And of course, form the host itself : > echo "10800 11211 47100 47500" | xargs -n 1 nc -vz 10.17.10.56 > Connection to 10.17.10.56 10800 port [tcp/*] succeeded! > Connection to 10.17.10.56 11211 port [tcp/*] succeeded! > Connection to 10.17.10.56 47100 port [tcp/*] succeeded! > Connection to 10.17.10.56 47500 port [tcp/*] succeeded! echo "10800 11211 > 47100 47500" | xargs -n 1 nc -vz 10.17.10.55 > Connection to 10.17.10.55 10800 port [tcp/*] succeeded! > Connection to 10.17.10.55 11211 port [tcp/*] succeeded! > Connection to 10.17.10.55 47100 port [tcp/*] succeeded! > Connection to 10.17.10.55 47500 port [tcp/*] succeeded! Without adding the > forced IP and the BasicAddressResolver I had the same results. Regards > > Le 07-Dec-2022 11:03:55 +0100, [email protected] a crit: > Thanks for the clarification. It seems like your containers could not see > each other via network. Could you please perform "cross check" out of the > container? Say, you enter to the container with ip 10.17.10.55 and try to > perform check to 10.17.10.56. You can use 'docker exec -it bash' to enter > to the container. If there is no connection between containers, try to > configure the docker network for therm. More info here > https://docs.docker.com/network/network-tutorial-standalone/ -- Best > regards, Aleksandr > > On 2022/12/06 08:27:20 [email protected] wrote: > > Hi, Here is the whole configuration file with my latest experiements : > > > > 10.17.10.55 > > 10.17.10.56 > > > > 10.17.10.55 and 56 are the external addresses of the the hosts. Of course > > each node as a different interfal IP address and ConsistendId. Regards > > > > Le 05-Dec-2022 12:59:39 +0100, [email protected] a crit: > > Hi, could you please share the TCP/IP Discovery configuration? If you have > > not configured it please check this > > https://ignite.apache.org/docs/latest/clustering/tcp-ip-discovery -- Best > > regards, Aleksandr > > > > On 2022/11/25 11:27:19 [email protected] wrote: > > > Hi, I am trying to setup a two nodes replicated cluster, in an > > > active/passive way. On each node, a Java webapp will be accessing the > > > local database instance using JDBC. The app itself is stateless, but I > > > need synced datas from the database. Ignite seems to be a nice choice, > > > as I don't need advanced SQL features but replication and simplicity. So > > > I am running an Ignite container on two differents hosts (10.17.10.55 and > > > 10.17.10.56), using : docker run -v > > > "/tmp/ignite.xml:/opt/ignite/apache-ignite/config/default-config.xml" -p > > > "10800:10800" -p "11211:11211" -p "47100:47100" -p "47500:47500" -p > > > "49112:49112" apacheignite/ignite:latest And here is the content of the > > > file "ignite.xml" (of course, "consistentId" value is different for each > > > host, the rest is the same). > > > > > > 10.17.10.55 > > > 10.17.10.56 > > > > > > As far as I can see, when the containers are up, ports are exposed and > > > firewall does not block communications (I have runned these two commands > > > on both hosts, in order to "cross check"): echo "10800 11211 47100 > > > 47500" | xargs nc -vz 10.17.10.55 > > > Connection to 10.17.10.55 10800 port [tcp/*] succeeded! > > > Connection to 10.17.10.55 11211 port [tcp/*] succeeded! > > > Connection to 10.17.10.55 47100 port [tcp/*] succeeded! > > > Connection to 10.17.10.55 47500 port [tcp/*] succeeded! echo "10800 > > > 11211 47100 47500" | xargs nc -vz 10.17.10.56 > > > Connection to 10.17.10.56 10800 port [tcp/*] succeeded! > > > Connection to 10.17.10.56 11211 port [tcp/*] succeeded! > > > Connection to 10.17.10.56 47100 port [tcp/*] succeeded! > > > Connection to 10.17.10.56 47500 port [tcp/*] succeeded! But the two > > > Ignite instances don't seem to communicate, as I get an error message > > > like : Failed to connect to any address from IP finder (make sure IP > > > finder addresses are correct and firewalls are disabled on all host > > > machines): [/10.17.10.55:47500, /10.17.10.56:47500] Did I miss something > > > ? A quick and dirty try with a local docker-compose and a config file > > > which is the base of the one I tried here was running flawlessly. > > > Regards > > > > > > ------------------------------------------------------------------------------------------------- > > > FreeMail powered by mail.fr > > > > > ------------------------------------------------------------------------------------------------- > > FreeMail powered by mail.fr > > > > ------------------------------------------------------------------------------------------------- > > FreeMail powered by mail.fr > > > > ------------------------------------------------------------------------------------------------- > FreeMail powered by mail.fr > ------------------------------------------------------------------------------------------------- FreeMail powered by mail.fr
