Problem Statement/Rationale

Yesterday I started a long-running deleteMany operation against a large collection (it was deleting about 100GB of data. At some point I got tired of waiting for it and pressed CTRL+C.

The shell correctly returned and printed "Interrupting operation on the server", however I then realised that the operation continued to run and it showed-up in db.currentOp.

When I found the operation in db.currentOp I ran db.killOp and it correctly got interrupted.

I did not think to include any logs from the server and unfortunately I shut down the server. But I am hoping that it should be easily reproducible, by just inserting some large amount of data and then running a long deleteMany.

Steps to Reproduce

Inserting some large amount of data and then running a long deleteMany. I am not certain it is problem specifically to deleteMany, but this is the one long-running operation that I could think of. Most likely a multi-update will have the same problem.

Expected Results

Operation is interrupted.

Actual Results

Operation was not interrupted.