You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
At some point we would like to support GCS in the same way we currently support S3 as a backup storage.
How could this be improved?
Similar to what was done with alpakka's s3 client, functionality needs to be upstreamed to Alpakka GCS client that allows us to both get the uploaded parts (aka chunks) of a currently incompletely multipart upload ergo akka/alpakka#2730 and also the ability to resume the upload of a file while providing context (which in our case is Kafka cursors) ergo akka/alpakka#2770
One thing to consider is that GCS appears to have 2 API's. Their original JSON based API which is what the current Alpakka' GCS client implements and also an S3 compatible API which they added later on. It needs to be investigated whether GCS's JSON based API even supports the functionality we need. If not the changes required would be more extensive as we would need to either re-implement the current GCS client using S3's API or whether we should make a completely new GCS client with the current on in tact (also need to consider if we should share code between Alpakka's S3 client and GCS's S3 client).
Is this a feature you would work on yourself?
I plan to open a pull request for this feature
The text was updated successfully, but these errors were encountered:
What is currently missing?
At some point we would like to support GCS in the same way we currently support S3 as a backup storage.
How could this be improved?
Similar to what was done with alpakka's s3 client, functionality needs to be upstreamed to Alpakka GCS client that allows us to both get the uploaded parts (aka chunks) of a currently incompletely multipart upload ergo akka/alpakka#2730 and also the ability to resume the upload of a file while providing context (which in our case is Kafka cursors) ergo akka/alpakka#2770
One thing to consider is that GCS appears to have 2 API's. Their original JSON based API which is what the current Alpakka' GCS client implements and also an S3 compatible API which they added later on. It needs to be investigated whether GCS's JSON based API even supports the functionality we need. If not the changes required would be more extensive as we would need to either re-implement the current GCS client using S3's API or whether we should make a completely new GCS client with the current on in tact (also need to consider if we should share code between Alpakka's S3 client and GCS's S3 client).
Is this a feature you would work on yourself?
The text was updated successfully, but these errors were encountered: