Public bug reported: According to AWS docs, using newer instance types that use NVME drivers (c5/m5), the io timeout should be set to maximum (ideally 4294967295). https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/nvme-ebs-volumes.html
It appears this is done for 16.04, but in 18.04 it is the default. Test done by spinning up an m5.large instance in ap-southeast-2. Ubuntu 16.04 AMI on AWS (latest AMI - 20180126): $ cat /sys/module/nvme/parameters/io_timeout 4294967295 $ uname -r 4.4.0-1049-aws Ubuntu 18.04 AMI on AWS (latest nightly AMI - 20180323): $ cat /sys/module/nvme_core/parameters/io_timeout 255 $ uname -r 4.15.0-1001-aws Perhaps as part of the move to nvme_core this got lost. ** Affects: linux (Ubuntu) Importance: Undecided Status: New ** Summary changed: - default io_timeout for nvme is 255 on AWS + bionic: default io_timeout for nvme is 255 on AWS -- You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. https://bugs.launchpad.net/bugs/1758466 Title: bionic: default io_timeout for nvme is 255 on AWS To manage notifications about this bug go to: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1758466/+subscriptions -- ubuntu-bugs mailing list ubuntu-bugs@lists.ubuntu.com https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs