Linux: Mounting an NFS Share

JaredBusch

@scottalanmiller said in Linux: Mounting an NFS Share:

@JaredBusch said in Linux: Mounting an NFS Share:

I have always wished that the mount command had to switch to put the damn thing Into fstab.

You mean like --perm and it just appended into fstab? That would rock.

Yeah pretty much anything like that

AdamF

Can you explain why if I just mount an NFS share, using the command below, I can get a data transfer/throughput speed of approx 300 MB/s

mount 192.168.2.2:/var/nfs_share1 /mnt/nfs1

BUT,

If I put an entry in fstab, it dramatically slows the transfer to about 35-40 MB/s

192.168.2.2:/var/nfs_share1 /mnt/nfs1 nfs rw,sync,hard,intr 0 0

What causes that?

scottalanmiller

@fuznutz04 said in Linux: Mounting an NFS Share:

Can you explain why if I just mount an NFS share, using the command below, I can get a data transfer/throughput speed of approx 300 MB/s
mount 192.168.2.2:/var/nfs_share1 /mnt/nfs1
BUT,

If I put an entry in fstab, it dramatically slows the transfer to about 35-40 MB/s
192.168.2.2:/var/nfs_share1 /mnt/nfs1 nfs rw,sync,hard,intr 0 0
What causes that?

sync

scottalanmiller

"The default export behavior for both NFS Version 2 and Version 3 protocols, used by exportfs in nfs-utils versions prior to nfs-utils-1.0.1 is "asynchronous". This default permits the server to reply to client requests as soon as it has processed the request and handed it off to the local file system, without waiting for the data to be written to stable storage. This is indicated by the async option denoted in the server's export list. It yields better performance at the cost of possible data corruption if the server reboots while still holding unwritten data and/or metadata in its caches. This possible data corruption is not detectable at the time of occurrence, since the async option instructs the server to lie to the client, telling the client that all data has indeed been written to the stable storage, regardless of the protocol used."

http://nfs.sourceforge.net/nfs-howto/ar01s05.html

scottalanmiller

So when you mount with the mount command without specifying, you get async. When you put it into fstab, you manually overrode the async to go with sync. So a major change.

krisleslie

@scottalanmiller are the nfs shares faster than Windows SMB shares?

AdamF

@scottalanmiller so safer, but slower essentially.

DustinB3403

@fuznutz04 said in Linux: Mounting an NFS Share:

@scottalanmiller so safer, but slower essentially.

Slower usually is safer.

scottalanmiller

@krisleslie said in Linux: Mounting an NFS Share:

@scottalanmiller are the nfs shares faster than Windows SMB shares?

NFS is generally faster than SMB. SMB is traditionally super slow, NFS is screaming fast. SMB has improved over time, but NFS typically beats it.

scottalanmiller

@fuznutz04 said in Linux: Mounting an NFS Share:

@scottalanmiller so safer, but slower essentially.

Sync is safer for writes, but yeah, way slower.

AdamF

@scottalanmiller Do you typically use the Sync then as a best practice, even though it is much slower to ensure there is no corruption?

scottalanmiller

@fuznutz04 said in Linux: Mounting an NFS Share:

@scottalanmiller Do you typically use the Sync then as a best practice, even though it is much slower to ensure there is no corruption?

Totally depends on the need. A database, hell yeah sync that. A general document file server? Nah, async is fine.

AdamF

@scottalanmiller Sounds solid. The speed difference is dramatic. Thanks.