MSP field troubleshooting
Practical troubleshooting paths for real-world support failures.
Start with the symptom, confirm scope, isolate the layer, apply the safest evidence-backed fix, and document the result. This toolkit is built for the middle of a ticket: when the user needs an answer, the dashboard is only half useful, and the next step has to be evidence-based.
Solution Atlas
Grouped by the way failures actually travel.
The library is organized around domains, but the visual system keeps the operator's real question in view: which layer is failing, which adjacent layer still works, and what evidence proves it?
Layer Isolation
Move from symptom to layer before applying a fix. Identity, policy, network, client, service, data, and dependency failures leave different evidence.
Evidence Packet
Escalation gets cleaner when the packet already contains timestamps, IDs, traces, screenshots, affected scope, and what has been ruled out.
How to Use This Toolkit
- Start with the closest matching symptom, not the product name alone.
- Confirm whether the issue follows the user, device, network, tenant, site, or service.
- Work Tier 1 before Tier 2 unless there is clear outage impact or security exposure.
- Document exact errors, timestamps, screenshots, affected scope, and working comparisons.
- Escalate with evidence: logs, test results, affected objects, and what has already been ruled out.
Common MSP Fire Drills
- Outlook stuck in disconnected mode after password change
- Outlook prompts for password every morning on one laptop
- VPN connects but cannot reach internal file shares
- Printer installed but jobs stay in queue
- BitLocker recovery key prompt after firmware update
- Teams sign-in loop after MFA enrollment
- Windows updates failing with 0x800f errors
- Domain-joined PC says trust relationship failed
- MFA prompts delayed or never arriving
- Website contact form sends mail intermittently
- Syncro agent shows online but scripts never return output
Technician Rule of Thumb
If a fix is destructive, noisy, or hard to reverse, prove the layer first. Rebuild the Outlook profile after token and credential checks. Reinstall an RMM agent after service and script-history checks. Reboot a server after logs and service state explain why the reboot is justified.
Jump to a troubleshooting path
Use these quick paths when you already know the failure pattern and need to get to the right layer fast.