-
- Influencer
- Posts: 17
- Liked: 2 times
- Joined: Mar 06, 2023 3:53 pm
- Full Name: Brandon Halloran
- Contact:
NAS vs Transport repository protocols
Hello,
Kinda semantics oriented question.
I have a physical NAS server (HPE storeeasy) that is the target of a NAS file share repository. I was looking into using a USB as removable storage since my tape system replacements are back ordered. So I looked into how to add the USB to my NAS server (all other servers are virtual). To do that I had to add the NAS server as a managed windows server in the backup infrastructure. When doing the add direct attached for the USB, I noticed that I could see the volume where my NAS share resides.
So my question is. Is it better/faster to use the file share (NAS) repository protocols or is it faster/better using the "Direct attached storage" managed server transport protocols to reach the volume?
The rest of my veeam system is an all in one install on a virtual machine.
Thank you
Kinda semantics oriented question.
I have a physical NAS server (HPE storeeasy) that is the target of a NAS file share repository. I was looking into using a USB as removable storage since my tape system replacements are back ordered. So I looked into how to add the USB to my NAS server (all other servers are virtual). To do that I had to add the NAS server as a managed windows server in the backup infrastructure. When doing the add direct attached for the USB, I noticed that I could see the volume where my NAS share resides.
So my question is. Is it better/faster to use the file share (NAS) repository protocols or is it faster/better using the "Direct attached storage" managed server transport protocols to reach the volume?
The rest of my veeam system is an all in one install on a virtual machine.
Thank you
-
- Veeam Software
- Posts: 3812
- Liked: 643 times
- Joined: Aug 28, 2013 8:23 am
- Full Name: Petr Makarov
- Location: Prague, Czech Republic
- Contact:
Re: NAS vs Transport repository protocols
Hi Brandon and Welcome to Veeam R&D Forums!
I'd expect better performance with the directly attached storage even if the managed server itself is virtual. I guess you should use this option as well.
Thanks!
I'd expect better performance with the directly attached storage even if the managed server itself is virtual. I guess you should use this option as well.
Thanks!
-
- Influencer
- Posts: 17
- Liked: 2 times
- Joined: Mar 06, 2023 3:53 pm
- Full Name: Brandon Halloran
- Contact:
Re: NAS vs Transport repository protocols
Thank you,
I will try to use the rotated storage option for the USB.
I will give the direct attached storage option a try for my non-rotated drive backups as well.
Thank you,
Brandon
I will try to use the rotated storage option for the USB.
I will give the direct attached storage option a try for my non-rotated drive backups as well.
Thank you,
Brandon
-
- Influencer
- Posts: 17
- Liked: 2 times
- Joined: Mar 06, 2023 3:53 pm
- Full Name: Brandon Halloran
- Contact:
Re: NAS vs Transport repository protocols
Just in case anyone was curious
Did a test of a single VM full backup using both protocols (Not doing USB copy backup yet) to the same server, same disk volume, same proxy. To eliminate differences as much as possible.
VM1 backup to NAS repository on NASserver2 to volume E through Proxy VeeamP1 (separate server) - 4 minutes 30 seconds - processing rate 161 MB/s - throughput 250MB/s
VM1 backup to DAS repository on NASserver2 to volume E through Proxy VeeamP1 (separate server) - 2 minutes 40 seconds - processing rate 897 MB/s - throughput 444MB/s
I guess the NAS protocol is either slower or is throttled (results were consistent running test 3 times)
Did a test of a single VM full backup using both protocols (Not doing USB copy backup yet) to the same server, same disk volume, same proxy. To eliminate differences as much as possible.
VM1 backup to NAS repository on NASserver2 to volume E through Proxy VeeamP1 (separate server) - 4 minutes 30 seconds - processing rate 161 MB/s - throughput 250MB/s
VM1 backup to DAS repository on NASserver2 to volume E through Proxy VeeamP1 (separate server) - 2 minutes 40 seconds - processing rate 897 MB/s - throughput 444MB/s
I guess the NAS protocol is either slower or is throttled (results were consistent running test 3 times)
-
- VP, Product Management
- Posts: 7200
- Liked: 1547 times
- Joined: May 04, 2011 8:36 am
- Full Name: Andreas Neufert
- Location: Germany
- Contact:
Re: NAS vs Transport repository protocols
Not sure if I can follow completely.
Are you saying with the first sentence:
VM1 backup to NAS repository on NASserver2 to volume E through Proxy VeeamP1 (separate server) - 4 minutes 30 seconds - processing rate 161 MB/s - throughput 250MB/s
That you read with VeeamP1 go over network to NASserver2 (selected as Veeam Gateway server) and write to a share (SMB) on the same NASserver2?
If you have configured it like this, then yes, there is natural overhead in the processing as you send the data 3 times through the operating system network stack of the NAS server 2 and additionally as well through a file server, instead of writing it directly to disk as you did with your second scenario.
Are you saying with the first sentence:
VM1 backup to NAS repository on NASserver2 to volume E through Proxy VeeamP1 (separate server) - 4 minutes 30 seconds - processing rate 161 MB/s - throughput 250MB/s
That you read with VeeamP1 go over network to NASserver2 (selected as Veeam Gateway server) and write to a share (SMB) on the same NASserver2?
If you have configured it like this, then yes, there is natural overhead in the processing as you send the data 3 times through the operating system network stack of the NAS server 2 and additionally as well through a file server, instead of writing it directly to disk as you did with your second scenario.
Who is online
Users browsing this forum: Amazon [Bot], neil.henderson, Semrush [Bot] and 151 guests