L2 Support Engineer · Fintech · Week 1
Week 1 Day 3
Today's Topic

Jira & SLA

Every problem a client reports becomes a ticket. You live inside tickets as an L2 engineer. Today you learn how tickets are born, how they move, and how fast you need to respond.

Ticket Workflow Severity Levels Escalation SLA
01 The Simple Idea First
Real-life Analogy

Imagine you go to a hospital. You walk in and tell the receptionist your problem. They write it down, give you a token number, and a doctor is assigned to you. If your issue is really serious (like a heart attack) you skip the queue. If it's a small cut, you wait your turn.

Jira is that hospital system. Every client complaint = a ticket. Every ticket has a number, a status, and someone assigned to fix it. SLA is the rule that says "how fast must the doctor respond." A heart attack = respond in 2 minutes. A small cut = respond in 2 hours.

What is Jira?

Jira is a tool companies use to track problems and tasks. When a client reports an issue — like "payments are failing" — someone creates a Jira ticket. That ticket tracks everything: what the problem is, who is working on it, what has been done, and whether it's fixed.

As an L2 support engineer, Jira is where your work lives. You'll open it every morning, see what tickets are assigned to you, and work through them one by one.

What is SLA?

SLA = Service Level Agreement. It's basically a promise your company makes to the client: "We will respond to your problem within X hours." If you miss that promise, it's a breach — and that's a big deal in fintech. Clients can penalize your company for it.

Different severity levels have different SLA times. A system-down emergency has a much tighter SLA than a minor UI bug.

02 The Ticket Lifecycle

How a ticket moves from start to finish

📥
Open
Ticket is created. Problem is reported. Nobody has touched it yet.
⚙️
In Progress
An engineer picks it up and starts investigating the issue.
🔍
In Review
A fix is ready. Someone else checks it to make sure it actually works.
Resolved
Fix confirmed. Client is informed that the issue has been fixed.
🔒
Closed
Client confirms everything is OK. Ticket is officially done and archived.

One important rule about tickets

A ticket should never just sit there silently. Even if you don't have a fix yet, you must update the ticket with what you've tried, what you found, and what you're doing next. This is called adding comments to the ticket.

The client can see the ticket status. If it stays "Open" for 3 hours with no update, they get nervous. Keep it moving — even a comment like "Investigating, found the issue is in the Payments module" is better than silence.

03 Severity Levels — How Bad Is It?
🔴 P1
Critical
Respond: 30 min · Fix: 2 hrs

Everything is down. No one can use the system. Real money is being affected right now. All hands on deck. Wake people up if needed.

🟠 P2
High
Respond: 2 hrs · Fix: 8 hrs

A major feature is broken but the system is still partly working. For example, payments work but refunds are failing for all users.

🟡 P3
Medium
Respond: 4 hrs · Fix: 24 hrs

Something is broken but it affects only some users or a non-critical feature. There's usually a workaround available.

🟢 P4
Low
Respond: 8 hrs · Fix: 72 hrs

Minor issues — a wrong label, a UI glitch, a small report formatting error. No one's workflow is blocked.

⏱️ SLA Quick Reference
Priority Example in Fintech First Response Resolution Target Who Gets Notified
P1 — Critical All payments down, core banking offline 30 minutes 2 hours L2 + L3 + Manager + Client
P2 — High Refunds failing, login broken for many users 2 hours 8 hours L2 + L3
P3 — Medium Reports not generating, slow dashboard 4 hours 24 hours L2 Engineer
P4 — Low Wrong date format, minor UI bug 8 hours 72 hours L2 Engineer
04 Escalation — When You Can't Fix It Alone
🚨 Escalation Path — Who Handles What
L1 Support

First contact. Picks up the ticket, asks basic questions, tries simple fixes. If they can't solve it in the agreed time, they escalate to L2. (That's where you come in.)

L2 Support ← You

Deeper investigation. You dig into logs, check the database, test in UAT, try to reproduce the issue. You know the product well. If the fix needs a code change or is beyond your access, you escalate to L3.

L3 / Dev Team

Code-level fixes. Developers or senior engineers who can actually change the source code. Only get involved when L2 confirms the problem is a real bug — not a config issue or user error.

Golden Rule of Escalation

Never escalate a ticket without first doing your homework. When you pass a ticket to L3, you must include: what the problem is, what you already tried, what you found in the logs, and why you believe it needs a code fix.

Sending a ticket with just "client is complaining, please check" is a bad escalation. It wastes everyone's time and makes you look lazy.

05 Dummy Ticket — Hands-on Lab

How to fill a proper Jira ticket

Every ticket must have: a clear title, the client name, which product/module is affected, severity level, steps to reproduce the issue, and what you've done so far. Here's a real example of what a good ticket looks like:

FIN-1042 In Progress P1 — Critical Opened: Today 09:14 AM
Alpha Bank — Payment Gateway completely down, all transactions failing
Client Alpha Bank
Product Payment Gateway
Module Transaction Engine
Environment PROD
Assigned To You (L2)
SLA Deadline 11:14 AM (2 hrs)
Client reports that all payment transactions are failing since 09:00 AM. Error message shown to end users: "Transaction could not be processed. Please try again." No transactions have gone through in the last 14 minutes. Estimated 500+ users affected. Client's operations team has escalated to their management.
Steps Taken So Far (L2 Notes)
01Checked server status — Application server is running, no CPU spikes.
02Checked application logs — Found repeated error: DB_CONNECTION_TIMEOUT starting at 09:01 AM.
03Checked database — Found the DB connection pool is exhausted. Too many open connections, no new ones allowed.
04Escalating to L3 Dev team to increase connection pool limit and investigate root cause of connection leak.
FIN-1051 Open P3 — Medium Opened: Today 11:30 AM
Gamma Finance — Monthly loan report not generating for October
Client Gamma Finance
Product Core Banking
Module Reporting Module
Environment PROD
Assigned To You (L2)
SLA Deadline Tomorrow 11:30 AM (24 hrs)
Client tried to generate the October loan summary report today. The report job runs but produces an empty file. No error shown to the user. Previous months (Aug, Sep) worked fine. Client needs the report for a board meeting tomorrow.
Next Steps (Your Investigation Plan)
01Check report scheduler logs to see if the job ran and at what time.
02Check if October loan data actually exists in the database.
03Compare report config for October vs September to spot any difference.
04Update ticket with findings and inform client within 4 hours.
06 Real L2 Scenarios
01

A ticket comes in at 2 PM marked P1. Your SLA says you must respond in 30 minutes. You stop everything else and jump on this first. Even if you have 5 other tickets open — P1 is always the priority.

02

You've been working on a ticket for 2 hours and still can't find the fix. Don't sit on it silently. Update the ticket with what you've tried, change status to "Escalated", and send it to L3 with full notes. SLA clock is still running.

03

A client calls and says "this is urgent, please make it P1." But it's just a report formatting issue. You assess the actual business impact — is any money blocked? Are users unable to work? If not, it stays P3. Severity is based on impact, not the client's mood.

04

You fixed an issue and moved the ticket to "Resolved." The client replies: "Actually it's still broken on our end." — You reopen the ticket, change status back to "In Progress", and investigate again. Never close a ticket without client confirmation.

✅ Day 3 Outcomes — Can You Do This?