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
The archives folder contains a 12GB file, and accessing it a few times through an application causes the memory usage of mount-s3 to increase. After 10 minutes, the following message is output to the messages log and the program terminates abnormally.
Sep 10 04:38:53 test-server systemd-coredump[15268]: Resource limits disable core dumping for process 12815 (mount-s3).
Sep 10 04:38:53 test-server systemd-coredump[15268]: Process 12815 (mount-s3) of user 0 dumped core.
The EC2 instance type is c6in.2xlarge, with 16GB memory and 10GB swap.
The memory increase rate is so high, is there any way or measure to reduce it?]
Hi @mura890, thanks for sharing your use case. We are aware about the issue and are actively working on a fix for that in #987, but can not provide any dates for resolution just yet. Unfortunately, with current Mountpoint version there is no way of limiting the memory consumption apart from adjusting the workload (reducing the number of concurrently open files and the reading speed).
Regarding your case, from the logs I can see that you are having a non-containerised read workload, with peak memory consumption at around 13.5GiB. The memory usage is dominated by prefetch queue (reported by prefetch.bytes_in_queue metric), which is exactly the case we're focused on.
Please, use #987 to track resolution and let us know if you have any further questions. We may close this issue as a duplicate in a while.
I forgot to mention --maximum-throughput-gbps CLI option of Mountpoint, which may be used to reduce the memory consumption. It's actual effect depends on the workload pattern, but it is definitely an option to explore. Generally, the lower value you set to it, the lower memory consumption you should expect.
Mountpoint for Amazon S3 version
mount-s3 version:1.7.2
AWS Region
ap-northeast-1
Describe the running environment
I am running an S3 bucket in the same account mounted on EC2 (OS: Rocky Linux 8.10)
Mountpoint options
What happened?
I created a service file for mount-s3 as follows, registered it as a service, and made it work.
/lib/systemd/system/mount-s3.service
The archives folder contains a 12GB file, and accessing it a few times through an application causes the memory usage of mount-s3 to increase. After 10 minutes, the following message is output to the messages log and the program terminates abnormally.
The EC2 instance type is c6in.2xlarge, with 16GB memory and 10GB swap.
The memory increase rate is so high, is there any way or measure to reduce it?]
Attach the log obtained with the "--debug --log-metrics" option.
mountpoint-s3-2024-09-10T04-26-12Z.log
Relevant log output
The text was updated successfully, but these errors were encountered: