A server leaves evidence.
When a website or service behaves unexpectedly, that evidence is often more useful than guessing.
You do not need to memorize every command before learning what to look for.
Start by understanding the major sources of information.
Access logs
A web-server access log records requests.
Depending on configuration, a log entry can include:
time
client IP address
HTTP method
requested URL
HTTP status
response size
referrer
User-Agent
Access logs can answer questions such as:
Was this URL requested?
When?
How often?
What status did the server return?
Was the requester a browser, crawler or other client?
Error logs
Error logs record problems reported by the web server or connected application environment.
Examples can include:
missing files
permission failures
application crashes
upstream failures
configuration errors
script errors
resource problems
When a 500 error appears in the browser, the error log is often one of the first places to inspect.
Application logs
Custom applications can maintain their own logs.
These may contain information the web server cannot know, such as:
failed jobs
database exceptions
user actions
API failures
import results
business-rule errors
The quality of an application’s logs can determine how difficult the application is to support.
Mail logs
Mail systems can record:
connections
authentication
delivery attempts
rejections
routing
bounces
spam decisions
A message “not showing up” is easier to investigate when the delivery path can be followed through logs.
Scheduled-job logs
Cron jobs and other scheduled processes should leave enough evidence to answer:
Did it start?
Did it finish?
How long did it run?
How much work did it process?
Did anything fail?
When did it last succeed?
Silent automation is difficult to trust.
CPU
CPU measurements show how much processor time the system is using.
High CPU is a symptom, not automatically the cause.
The next question is:
Which process is using it, and why?
Memory
Memory usage helps show whether applications have enough working space.
A server can behave badly when memory is exhausted or when excessive swapping occurs.
Again, identify the process rather than treating “high memory” as the complete diagnosis.
Disk space
A full disk can break:
databases
logs
uploads
sessions
temporary files
backups
applications
Checking available disk space is simple and often overlooked.
Processes
A process list shows what programs are running.
It can reveal:
stuck jobs
unexpected duplicate workers
high-resource processes
missing services
long-running commands
Network connections
Connection information can help determine whether a service is listening, whether applications are connecting outward and whether expected network relationships exist.
Time connects the evidence
Record when the problem occurred.
Then compare:
access logs
error logs
application logs
mail logs
system measurements
recent deployments
scheduled jobs
The timestamp can connect events that initially look unrelated.
Read first
Before changing configuration:
capture the error
record the time
identify the affected service
check the relevant logs
look for recent changes
identify the failing layer
Then make the smallest justified change.
Logs are not clutter.
They are part of the interface between the operator and the machine.
Put This Lesson to Work
Choose a real website or project you control. Identify where this lesson applies, record who is responsible for that part, and identify one change or check you can safely make.
Worked example
For a domain-email lesson, record the separate website and email providers. Look up the existing MX records without changing them, then confirm which provider receives mail.
Your finished result
A short explanation of the system, its owner, and a checked next action. Keep passwords and recovery codes out of your notes.
Check your understanding
- Can you explain the main idea without repeating a slogan?
- What evidence or example supports your explanation?
- Which decision will you make differently, and why?