Hi @akash7 ,
On Superdome 2, “I2C Buses: Failed” on an IOX generally means something on the internal I2C management chain is stuck or not responding. Because I2C is a shared bus, a single faulty device can pull the bus low and make the entire shelf report failed. Since you replaced the IOX 9 chassis but reused the existing I/O cards, the most likely causes are:
- A faulty or marginal PCIe I/O card in IOX 9 that’s holding the I2C lines
- A card with incompatible/old firmware causing I2C timeouts
- A seating/connector issue or contamination in a slot
- Less likely (but possible): a bad fan/FRU on the shelf or a backplane issue in the replacement unit, though other sensors look OK
Recommended isolation steps
Please schedule a brief maintenance window and do the following in order:
1) Soft reset the IOX shelf
- From the Superdome 2 OA/management interface, perform a reset of IOX 9’s management controller or power-cycle the IOX 9 shelf if there is a supported option. Sometimes an I2C bus lock clears with a controller reset.
- If there’s no software reset option, you can perform a cold power cycle by removing power to the shelf (both PSUs) for at least 60 seconds, then reapply power. Do this only during a window and after confirming redundancy and impact.
2) Reseat external items
- Verify and reseat the Xfabric cables to IOX 9 and the Xfabric modules on both ends. Your links show OK, but it’s quick to eliminate a marginal connection.
- Check both IOX 9 PSUs and fan trays: reseat them. Ensure both fans spin up and the LEDs look normal.
3) Check for obvious slot issues
- Inspect each IOX 9 PCIe slot for bent pins, debris, or damage.
- Ensure cards are fully seated and locking mechanisms engaged.
4) Remove all I/O cards from IOX 9 and retest
- Power down or offline the IOX shelf per HPE procedures (to avoid hot-plug impact on the management bus).
- Remove all PCIe cards from IOX 9.
- Power up IOX 9 and check if “I2C Buses” changes from Failed to OK.
- If I2C now reports OK with no cards installed: add cards back one at a time (or in small batches), checking the I2C status after each addition to identify the offending card/slot.
- If I2C still reports Failed with all cards removed: the replacement IOX shelf itself may have a backplane or management FRU issue. Proceed to step 7 and contact HPE.
5) Verify firmware baselines
- Ensure the complex firmware bundle (system/Onboard Administrator/Xfabric/IOX) is at a supported and consistent level.
- Check each I/O card’s firmware and option ROM versions against the Superdome 2 supported matrix for your exact model and OS. Update any cards that are behind. Even though “Firmware Mismatch: OK” is shown for the shelf, individual adapter firmware mismatches can still cause I2C issues.
6) Identify the culprit card
- When the I2C fails after reinserting cards, the last card/slot you reintroduced is the likely culprit. Swap that card into a different known-good slot or a different IOX to confirm the fault follows the card.
- If the fault stays with the slot, inspect that slot for damage and consider the IOX backplane as a suspect.
7) Collect logs and engage HPE if needed
- From the OA/complex manager, collect a full support/diag bundle (OA logs, hardware inventory, event logs). On Superdome 2, HPE may ask for an “sx3000/complex” log pack and IOX logs.
- Note the exact IOX 9 FRU part numbers and firmware, plus the list of cards/slots and their firmware versions.
- Open a case with HPE if:
- I2C still fails with all cards removed, or
- The issue follows the replacement shelf, or
- You identify a specific card model repeatedly causing I2C failures even with current firmware.
Regards,
Rakshitha K
I work at HPE
HPE Support Center offers support for your HPE services and products when and how you need it. Get started with HPE Support Center today.
[Any personal opinions expressed are mine, and not official statements on behalf of Hewlett Packard Enterprise]
