Building a highly available load balancer with keepalived and HAProxy

[@puneetloya1531]

Awesome explanation with crystal clear examples. Thank you!!

[@LinuxCloudHacks]

Glad you enjoyed it! Please stay tuned as there are more videos to come!

[@lawmaina78]

I spent whole morning researching how to do these stuffs until I came across this awesome video.

[@LinuxCloudHacks]

Great to hear! Good luck!

[@jocelyn-n-tech]

This guy is good. Very pleasant and clear explanations combined with examples. Bookmarking his channel.

[@LinuxCloudHacks]

Thank you for those kind words! Please stay tuned for more interesting content. I have a 50+ list of topic I'd like to cover. Just need to find the time :) Cheers!

[@bharath1234561]

Very clear explanation…Thank you

[@LinuxCloudHacks]

Glad you liked it! Thanks

[@piotrd8983]

Brilliant and from Poland :) Thanks a lot.

[@LinuxCloudHacks]

Thank you Piotr! Stay tuned for future videos!

[@vonelfensenf8151]

Nice video! But as always this example just shows basic HTTP traffic. I wonder if the HA setup would also work with long living TCP streams. I know that haproxy can proxy TCP too but I wonder if the socket of the client will still be intact in case one node/process goes down.

[@LinuxCloudHacks]

Hi! Please look at 'lvs_sync_daemon' option of keepalived. It allows you to synchronize connection states between primary and backup load balancers (assuming you are doing IPVS). I'll try to find some time to make a video about it. In the meantime you can look at my IPVS video kzbin.info/www/bejne/equmc6NoeL2mj7Msi=ss-Qvwdi7objdk9_

[@imtiazali1628]

Hi thanks for such a nice explaination. How can we sync the config in future can you please advise. Thanks

[@LinuxCloudHacks]

If those are standalone servers/VMs I would recommend something simple like exchanging SSH keys for password-less authentication and then doing a rsync based on a cron schedule.