Aiven-Open / Aiven-Open/tiered-storage-for-apache-kafka

Too many byte copies during S3 upload

Offen
#128 0 Kommentare 0 Reaktionen 0 zugewiesene Personen Auf GitHub ansehen
Vorherrschende Sprache
Java
Sterne
239
Forks
59
PR-Merge-Kennzahlen
Keine gemergten PRs in 30 T.

Beschreibung

Currently there are three extra bytes copy phase when we upload log segments onto S3:

1. A byte copy from local file to a byte[] in S3ClientWrapper.uploadLogFile: https://github.com/aiven/tiered-storage-for-apache-kafka/blob/main/s3/src/main/java/io/aiven/kafka/tiered/storage/s3/S3ClientWrapper.java#L179
2. A compress and encryption phase in cryptoIOProvider.compressAndEncrypt
3. Another ByteBuffer copy in S3OutputSteram in https://github.com/aiven/tiered-storage-for-apache-kafka/blob/e0c5da2285359c496d2be0c3de3259b2320b9b09/s3/src/main/java/io/aiven/kafka/tiered/storage/s3/S3OutputStream.java#L81

Aiven is hosting for other companies and so it needs to do the data encryption. But in our use case, our data and S3 bucket are only accessible to our company, there is no need to do so many copying and encryption. Can we have an option to skip all those steps?

Beitragsleitfaden

Beitragsleitfaden öffnen

Bewertung

Dieses Issue wurde noch nicht bewertet.

Neue Issues direkt in Ihr Postfach

Eine kurze Übersicht über anfängerfreundliche GitHub-Issues.