So sánh nhà cung cấp

Khác biệt cần lưu ý: An Azure block blob reaches about 190.7 TiB (50,000 blocks x 4,000 MiB), far above S3's 50 TB, but a single Put Blob write is capped at 5,000 MiB against S3's 5 GB single PUT — so Azure holds larger objects while both force chunking at a similar per-request size.

Khác biệt cần lưu ý: A single Cloud Storage object caps at 5 TiB, whereas an S3 object uploaded via multipart reaches 50 TB — roughly a 10x difference that decides whether large media masters or database snapshots can be stored as one object at all.

Khác biệt cần lưu ý: Cloud Storage IAM changes are only eventually consistent — granting or revoking access typically takes about a minute to take effect, as does recreating a deleted bucket. Object read-after-write itself is strongly consistent, so the surprise lands on permission changes, not data.

Khác biệt cần lưu ý: Azure forces an up-front choice of blob type that S3 has no equivalent of: block blobs reach about 190.7 TiB, append blobs only about 195 GiB, and page blobs 8 TiB. Picking append-blob for a log sink that later needs to hold terabytes is a migration, not a setting change.

Khác biệt cần lưu ý: Cloud Run allows a request timeout up to 60 minutes (3,600 seconds) against Lambda's hard 15-minute (900 second) ceiling, so a long-running transcode or batch job that must be split into steps on Lambda often runs unchanged on Cloud Run.

Khác biệt cần lưu ý: Cloud Run's request timeout reaches 60 minutes (3,600 seconds) against Lambda's hard 900-second ceiling, and it deploys a container image rather than a zip of function code — so work that Lambda forces you to split into orchestrated steps often stays a single Cloud Run request.