Add udev rule to increase read_ahead_kb when an NFS share is mounted

[hailmo-amzn]

Description of changes:
Adding a udev rule to increase the read_ahead_kb value to an old kernel default value when an nfs share is mounted to remedy latency seen with efs.

Before Change:

$ uname -r;sudo mount -t nfs4 -o nfsvers=4.1,hard,timeo=600,retrans=2,noresvport fs-b36e9c87.efs.eu-west-1.amazonaws.com:/ efs;sudo ./nfs-readahead show /home/ec2-user/efs/; time rsync  efs/test .
5.4.0-1.132.64.amzn2.x86_64
/home/ec2-user/efs 0:40 /sys/class/bdi/0:40/read_ahead_kb = 128

real    0m14.428s
user    0m3.842s
sys    0m1.099s

After Change:

$ uname -r;sudo mount -t nfs4 -o nfsvers=4.1,hard,timeo=600,retrans=2,noresvport fs-b36e9c87.efs.eu-west-1.amazonaws.com:/ efs;sudo ./nfs-readahead show /home/ec2-user/efs/; time rsync  efs/test .
5.4.0-1.132.64.amzn2.x86_64
/home/ec2-user/efs 0:40 /sys/class/bdi/0:40/read_ahead_kb = 15360

real    0m3.579s
user    0m3.848s
sys    0m0.929s

By submitting this pull request, I confirm that you can use, modify, copy, and redistribute this contribution, under the terms of your choice.

[nmeyerhans]

Committed as 9b1ff7d