Reading ASIC logs and preparing a support report
Turn an error message into a reproducible incident with timestamps, context and a clear next step.
ASIC.tools · Reviewed
Capture evidence before restarting
Write down when the problem began and whether it affected one device, one rack or the whole site. Export the available current and history logs using the supported interface before a restart removes useful context. Save the firmware version and the visible hardware status with the logs. Screenshots can support a report, but searchable text helps technicians locate repeated events and compare their sequence.
Read the sequence, not one line
Look for the first relevant warning before the final shutdown message. A later communication error may be a consequence of a power or temperature fault rather than its cause. Separate normal startup detection from failures during steady operation. Check whether the device clock is correct and note its time zone. Misaligned timestamps can make unrelated network and electrical events appear to happen together.
Group the symptoms
Separate connection and authentication problems from board detection, sensor, fan and power supply faults. Use the exact model’s documentation to interpret identifiers. Similar text can have different meanings across firmware versions. Note how many boards were expected and how many were detected. Do not turn a single error code into a confident diagnosis of a specific failed chip without the required service measurements.
Test one reversible change
Choose a low-risk test that matches the evidence, such as a known-good network cable for repeated link loss. Record the initial state, the change and the observation period. Electrical measurements inside a power supply and board-level repairs belong to trained service personnel. Repeatedly changing firmware, pool and power profile at once destroys the comparison and can make a recoverable problem harder to isolate.
Create a concise service package
Include model and serial identifier, firmware, fault onset, relevant log excerpt, operating conditions and tests already performed. Remove passwords, access tokens and unrelated personal data before sending it through the official support channel. State whether the fault is repeatable and whether the device is safely powered off. Keep an unchanged copy locally so later repairs can be compared with the original incident rather than memory.
