Skip to content
View tobiabidoye's full-sized avatar

Highlights

  • Pro

Block or report tobiabidoye

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
tobiabidoye/README.md

Hi, I'm Tobi

Software engineer at JPMorgan Chase (NYC), working on cloud HPC infrastructure. CS grad from UMBC. I like distributed systems, consensus, and infrastructure.

Languages: Go · C++ · Java · Python Tools: Terraform · Docker · AWS · PostgreSQL · Git

Open source

etcd (CNCF, Go, 52k+ stars)

All my merged PRs →

Merged PRs: #22542 · #22552 · #22554 · #22558 · #22583 · #22590 · #22591

  • Fixed an etcdctl txn panic when comparing leases (value parsed as a hex lease ID), with regression test TestParseCompare. Backported to release-3.6 (PR #22542).
  • Remediated OpenTelemetry vulnerability GO-2026-6505 across all three supported branches (3.5, 3.6, 3.7), one PR per branch, upgrading the tracing SDK and OTLP exporters to v1.45.0 without breaking release builds.
  • Tailored each backport to its branch's dependency constraints and iterated through maintainer review to approval on all 6 PRs.

Projects

Fault-tolerant, linearizable key-value store in Go on a from-scratch Raft implementation (leader election, log replication, persistence, snapshotting) with no external consensus or storage libraries.

  • Verified linearizability with Porcupine and failure-tested leader/follower crashes and snapshot catch-up.
  • Crash-safe durability via atomic write-then-rename with fsync; exactly-once client semantics through request deduplication.
  • Deployed as a 3-node AWS EC2 cluster across availability zones: 111 Put ops/sec, 143 ms p50, 290 ms p99.

MapReduce-style distributed prime counter in Go using gRPC and protobuf.

Full-stack on-campus food delivery app built for the UMBC community.

Find me

LinkedIn

Pinned Loading

  1. DistributedPrimeCounter DistributedPrimeCounter Public

    A distributed prime counter which uses a map reduce style pattern written in golang using protobuf and grpc.

    Go 1

  2. Distributed-Kv-Store Distributed-Kv-Store Public

    Open source distributed kv store using my own raft implementation. All written by hand. This is heavily based on my 6.5840 raft implementation.

    Go 1

  3. etcd-io/etcd etcd-io/etcd Public

    Distributed reliable key-value store for the most critical data of a distributed system

    Go 52.4k 10.6k