File tree Expand file tree Collapse file tree 2 files changed +10
-1
lines changed
terraform/modules/services/airflow Expand file tree Collapse file tree 2 files changed +10
-1
lines changed Original file line number Diff line number Diff line change @@ -24,12 +24,19 @@ def determine_latest_zarr(bucket: str, prefix: str):
2424 size_old += obj .size
2525
2626 # If the sizes are different, create a new latest.zarr
27+ s3hook .log .info (f"size_old={ size_old } , size_new={ size_new } " )
2728 if size_old != size_new and size_new > 500 * 1e3 : # Expecting at least 500KB
28- # Delete the old latest.zarr, if it exists
29+
30+ # delete latest.zarr
31+ s3hook .log .info (f"Deleting { prefix } /latest.zarr/" )
2932 if prefix + "/latest.zarr/" in prefixes :
3033 s3hook .log .debug (f"Deleting { prefix } /latest.zarr/" )
3134 keys_to_delete = s3hook .list_keys (bucket_name = bucket , prefix = prefix + "/latest.zarr/" )
3235 s3hook .delete_objects (bucket = bucket , keys = keys_to_delete )
36+
37+ # move latest zarr file to latest.zarr using s3 batch jobs
38+ s3hook .log .info (f"Creating { prefix } /latest.zarr/" )
39+
3340 # Copy the new latest.zarr
3441 s3hook .log .info (f"Copying { zarrs [0 ]} to { prefix } /latest.zarr/" )
3542 source_keys = s3hook .list_keys (bucket_name = bucket , prefix = zarrs [0 ])
@@ -40,6 +47,7 @@ def determine_latest_zarr(bucket: str, prefix: str):
4047 dest_bucket_name = bucket ,
4148 dest_bucket_key = prefix + "/latest.zarr/" + key .split (zarrs [0 ])[- 1 ],
4249 )
50+
4351 else :
4452 s3hook .log .info ("No changes to latest.zarr required" )
4553
Original file line number Diff line number Diff line change @@ -41,6 +41,7 @@ services:
4141 ECS_SECURITY_GROUP : $ECS_SECURITY_GROUP
4242 _AIRFLOW_WWW_USER_PASSWORD : ${PASSWORD}
4343 AIRFLOW_CONN_SLACK_API_DEFAULT : ${AIRFLOW_CONN_SLACK_API_DEFAULT}
44+
4445 user : " ${AIRFLOW_UID:-50000}:0"
4546 volumes :
4647 - data:/airflow
You can’t perform that action at this time.
0 commit comments