title | date | lastmod |
---|---|---|
Brag Doc |
2022-11-07 |
2024-07-12 |
As learnt from https://jvns.ca/blog/brag-documents/
Project: API auto failure detection, triaging and reporting internal tool
- Roll out for LATAM regions
- Build new service-specific email reporting flow to increase workflow efficiency and standardization for all service teams
- Helped to increase success rates from 60% to 90% in the UAT environment
Project: Lead development of Android EV charging payment application
- Stablisation of the application in the new Spectra POS hardware. Achieved 40% reduction in production-reported defects.
- Helped integrate a new payment platform "Soepay" with our application.
- Implemented abstracted Event Bus on top of event bus external library which could only cache a single item, in order to support multiple charging sessions
- Implemented and designed a power loss recovery flow. Used on device payment logs as a source of truth which then synchronised with the backend charging service to recover ongoing transactions. This allowed the team to be more confident in deploying for larger scale customers:
- Implemented a session expiry flow to capture payment when a user's charging session has gone on for too long:
- Goal: reduce overhead required to support multiple releases. Added github backport actions to help us push critical changes to past releases.
- Implemented a way to register and log desired event types to be sent to our backend servers. Goal: improve early visibility of indicators that could point to failures.
- Implemented a way to detect the current network status in order to perform logic such as retries.
- Debugged and fixed an OOM issue in a recursive network call:
- Goal: easier debugging of events. Refactored our charging session logic into a state machine:
- Debugged and fixed an issue with external HTTP library, increasing stability in poor network conditions (such as in basement carparks):
Improved diagnostic daemon features
- Install utility shell script for interfacing with the rasdaemon database when hhmd starts. Use script to enable DDR error count and add PCIE error count test for lightyear
- Enable eMMC bus width test and add hhm_get_emmc_bus_drive_strength and associated tests
- Include HPE Switchtec library for use to perform PCIe switch device id and firmware version checks as well as scratchpad R/W tests
- Add I2C PSU tests
- Add FPGA scratchpad rw test of the alternate MM
- Add unit tests increasing code line coverage by 40% Integrate Fan daemon features
- Add functions to check emc230x fan and controller configs to support testing fan controller misconfiguration recovery.
- Implement EEPROM reads for new device Debugged and fixed OOB which was causing temperature daemon to fail: Goal: improve debugging speed for LED Goal: optimize powerd. Profile time taken for main loop and interrupt thread , finding out that the time taken to complete the interrupt service was 9ms. Using a semaphore to signal to the main thread to handle the interrupt instead saved interrupt handler time to 2ms.