Lab catalog
77 labs · 11 tracks77 of 77loading…
Loading…
-
The service is denied access The knowledge base moved to a new disk, and the service no longer sees the articles — even though the permissions are fine and the service user reads them directly. Find what denies the access and give it back to the service without weakening its protection.Server security ~20 min not started
-
The config changes by itself The portal settings keep reverting to someone else's, even though they were just fixed. Put an audit rule on the file, find from the events who rewrites it and with what, and stop it without touching anyone else.Server security ~20 min not started
-
Container won't start A status page in a Docker container stopped opening after an edit. Find the cause and fix it.Containers: Docker and compose ~20 min not started
-
A service from a container The new version of the status page arrived as a Docker image. Run it in a container in place of the old service so that the page survives a docker restart and a machine reboot.Containers: Docker and compose ~20 min not started
-
A new image version The build of version 1.1 of the application fails. Build the app:1.1 image from the project directory and replace the running version 1.0 with it.Containers: Docker and compose ~20 min not started
-
Data after re-creation After the notes application was updated, the list is empty although the container is running. Find the notes and give them back to the application without losing any.Containers: Docker and compose ~20 min not started
-
A container restarts under load The import service moved into a container and has not processed a single export since — the container is up one moment and restarting the next. Find the cause and fix it so that the service handles the load within the limits it was given.Containers: Docker and compose ~25 min not started
-
The pricing service is down After a night-time restart the pricing service does not start. Its settings live under git, and several people have edited them. Bring back the working settings without losing yesterday's needed change and without rewriting the history.Automation: Git and Ansible ~20 min not started
-
The server does not pick up changes The developers pushed a change to the repository, but the server did not pick it up. Someone edited the settings by hand on the server last night, and that change is needed too. Update the working copy, keep both changes and push yours to the repository.Automation: Git and Ansible ~20 min not started
-
An unfinished playbook The shop status page is to be set up with an Ansible playbook instead of by hand. The playbook is started, but it fails and does not do everything. Finish it so that it sets the machine up by itself and a second run changes nothing.Automation: Git and Ansible ~25 min not started
-
First day on the server Sort out what the previous admin left behind: find the log, put the configs in place, count the failed deploys and get the tool back on the PATH.— ~30 min not started
-
A site and its database on two machines The orders site moved to node1 while its database stayed on node2, and the site stopped opening. Find every cause and leave the database open to the site only.— ~50 min not started
-
Cleaning up exports on a schedule The cleanup script run by cron deletes the wrong files, gets in the way of the backup and stays silent. Fix it and move it to a systemd timer with a log and a lock.— ~40 min not started
-
A compose stack after the move The orders service was moved from hand-made docker run commands to compose, and the stack won't come up. Find every cause and bring the service back with its existing orders.— ~40 min not started
-
A slow server: triaging three complaints The catalog site answers slowly, the server keeps running short of available memory, and under a rush of visitors requests fail with an error. Find the culprit behind each complaint, remove the causes so that they do not come back after a reboot, and write the triage down in a report.— ~45 min not started
-
A server ready for publication A new ticket server is about to go public, but the security review sent it back. Leave key-only login without root, close the extra ports, turn on brute-force protection and confine the service — so that all of it holds after a reboot.— ~60 min not started
-
A site on node2, by a playbook from node1 A colleague's playbook is meant to bring up a status page on node2 from node1, yet it finishes without errors and there is no page. Fix the project so that the playbook configures node2 and a second run changes nothing.— ~50 min not started
No lab on the topic you need? Suggest a lab