diff --git a/reports/whatokk_Lab9.pdf b/reports/whatokk_Lab9.pdf
new file mode 100644
index 0000000..7766328
Binary files /dev/null and b/reports/whatokk_Lab9.pdf differ
diff --git a/reports/whatokk_Lab9_experiment_record.md b/reports/whatokk_Lab9_experiment_record.md
new file mode 100644
index 0000000..e80c2ce
--- /dev/null
+++ b/reports/whatokk_Lab9_experiment_record.md
@@ -0,0 +1,16 @@
+# Lab9 Experiment Record
+
+- Experiment: Distributed Object Storage & Fault Tolerance Testing
+- Repository: https://github.com/whatokk/lab9-distributed-object-storage
+- Successful Actions run: https://github.com/whatokk/lab9-distributed-object-storage/actions/runs/27364710248
+- Execution environment: GitHub Actions Ubuntu runner
+- Storage simulation: MinIO with four mounted directories data1, data2, data3, data4
+- Bucket: my-cloud-storage
+- Test object: bigfile.bin, 20 MiB
+- Fault simulation: stopped cluster, deleted and recreated data3, restarted cluster
+- Result: object remained accessible; original and recovered files had identical SHA256 hash
+- Generated at: 2026-06-12 01:25:22 +08:00
+
+## Key Results
+
+MinIO distributed storage split object data and metadata across all four storage directories. After data3 was removed and recreated, the object `bigfile.bin` was still listed and readable. The SHA256 hashes of `bigfile.bin` and `recovered-bigfile.bin` matched exactly, verifying fault tolerance through erasure coding and distributed redundancy.