Skip to content

tp.md

Latest commit

 

History

History
66 lines (45 loc) · 1.97 KB

tp.md

File metadata and controls

66 lines (45 loc) · 1.97 KB
Raw

Load balancing and replication

During this practical work, we will play with load balancing, failover and replication.

Question 1

In this question you will configure a HA proxy installation.

In the question_1 folder you will find a docker-compose file. It contains 2 apache server that serves the same page. They are behind a HA proxy container. You will have the HA proxy configuration exposed.

  • Configure HA proxy in a round robin fashion
  • Configure HA proxy for stickyness

You can observe the results in the logs.

Question 2

In this question, you will set up shared IP mechanisms and experience failover.

You will need to launch virtualbox (as we need kernel modules for UCARP).

  • Create 3 machines
  • Configure them to communicate over the same (private) network with fixed IPs
  • Configure machine 1 and 2 to share an IP address
  • Ping the shared ip with machine 3
  • Shutdown machine 1
  • What do you observe ? How long did it took ?
  • Start machine 1 again.
  • Shutdown machine 2
  • Same question.

Question 3

In this question, we will set up a highly available NFS server.

  • Use the 3 previously created machines. We will rely on their shared IP.
  • Install a NFS server on machine 1 and 2
  • Create the device
  • Designate a master
  • Create the file system on the device.
  • Mount the file system on master
  • Export it over NFS on master (using shared IP)
  • Mount it using the client
  • Create a file
  • Shut down master
  • Declare the slave as primary
  • mount the volume and export it
  • re-connect the client if needed
  • See the file that was created

You can now automate the mouting process using UCARP

  ucarp-downscript  /usr/local/bin/vip-down
  ucarp-upscript    /usr/local/bin/vip-up

Miscaleneous

Useful links :