News for hackers. Stuff that matters.
Re:A shared file system for lambda functions by kleebeesh on Tuesday June 16, 2020 @06:01PM (#23544806)
I've found EFS enticing in theory but painfully slow and riddled with issues in practice. In the past I've tried it thinking "it's basically an EBS volume I can mount on > 1 EC2 instance," only to find terrible read performance and misc. low-level NFS errors.
Re:A shared file system for lambda functions by tidepod12 on Tuesday June 16, 2020 @06:57PM (#23545343)
Dunno your exact requirements or when you last tried it, but they did boost EFS's read speed (they claim by 400% [1]) as of this April, so it might be worth looking into again if you're still trying to find a solution.

1: https://aws.amazon.com/about-aws/whats-new/2020/04/amazon-el...

Re:A shared file system for lambda functions by spullara on Tuesday June 16, 2020 @07:10PM (#23545476)
That is great to know, thanks. It was pretty unusable the last time I tried it.
Re:A shared file system for lambda functions by VWWHFSfQ on Tuesday June 16, 2020 @10:52PM (#23547194)
Also insanely expensive.

Do you know how much it costs to write 3 MB/second to EFS? How much does it cost to read 26 MB/second from the same EFS volume?

That's $1,400/month.

Re:A shared file system for lambda functions by tw04 on Tuesday June 16, 2020 @07:14PM (#23545505)
You should try one of the various netapp options. Marketplace but significantly better performance and a platform that’s got several decades of nas experience.

https://cloud.netapp.com/cloud-volumes-service-for-aws

Whether amazon will ever let you mount it to lambda is a very different discussion.

Re:A shared file system for lambda functions by jrockway on Tuesday June 16, 2020 @11:13PM (#23547334)
I found it adequate to work around limitations of another system I was using. We were using a container coordinator thing (Convox) that couldn't attach EBS volumes, nor could it limit concurrent access to exactly one replica. So I used EFS which worked OK. I kept an eye on how we were using the burst credits, and picked a filesystem size that gave us enough IOPS. All in all, it worked fine (but I was perfectly happy to move off of it to EBS, of course).