Virtual environment data storage questions, seeking advice!
Hello fellow experts, I am seeking the advice of anyone who thinks they might be able to offer some help.
I work for a medium sized organization, roughly 200-250 users. We have implemented VMware a year ago and are now using Veeam to backup our VMs. Furthermore we have purchased a 2nd SAN for our virtual environment. We have 2 Hosts and about 18 VMs.
The problem we are foreseeing is as follows:
We have a datastore, lets call it "Shared drives datastore" that is 1.2TB, used for storage. This is where users store dept. data, shared data for all departments. This is almost filled up.
In addition to this we are now in the process of moving from roaming profiles to folder redirection/offline folders. This data resides on one of our last physical servers. We are going to move this data to the new SAN, we are allocating roughly 1.36TB to this specific datastore. The virtual disk will reside on the same VM that the other 1.2TB virtual disk resides.
We would like some way to ensure that we could expand that 1.36TB space to whatever we would please. We are going to be placing the folder redirected my docs, desktops and application data for over 200 users. In addition in order to reduce space on our mail server we are training users to delete emails and save important attachments in their my docs, thus further increasing data on the new datastore.
How do large enterprises cope with such massive amounts of data on VMs? Thank you for any advice, Marco.
I work for a medium sized organization, roughly 200-250 users. We have implemented VMware a year ago and are now using Veeam to backup our VMs. Furthermore we have purchased a 2nd SAN for our virtual environment. We have 2 Hosts and about 18 VMs.
The problem we are foreseeing is as follows:
We have a datastore, lets call it "Shared drives datastore" that is 1.2TB, used for storage. This is where users store dept. data, shared data for all departments. This is almost filled up.
In addition to this we are now in the process of moving from roaming profiles to folder redirection/offline folders. This data resides on one of our last physical servers. We are going to move this data to the new SAN, we are allocating roughly 1.36TB to this specific datastore. The virtual disk will reside on the same VM that the other 1.2TB virtual disk resides.
We would like some way to ensure that we could expand that 1.36TB space to whatever we would please. We are going to be placing the folder redirected my docs, desktops and application data for over 200 users. In addition in order to reduce space on our mail server we are training users to delete emails and save important attachments in their my docs, thus further increasing data on the new datastore.
How do large enterprises cope with such massive amounts of data on VMs? Thank you for any advice, Marco.
Andrew Hancock (VMware vExpert PRO / EE Fellow/British Beekeeper)
You can expand virtual disks, but your single virtual disks limit will always be 2TB, under VMware vSphere 4.x and 5.0, although the datastore can be greater than 2TB in VMware vSphere 5.0.
ASKER
Yes, I read that yesterday. I was hoping extending datastores would work but i read about the 2TB limit. Do large enterprises break up there data storage structures? Example: Marketing dept goes to virtual disk 1 with 750Gb allocated, Accounting goes to virtual disk 2 with 400Gb allocated and so on...
Creating GPOs for folder redirection PER department? Also allocating a virtual disk for each departments shared drives?
I see how this would be beneficial for backups as well. All Virtual disks would be smaller in size, in comparison to having one large 1TB+ virtual disks.
I would still like to hear more input from anyone wiling to offer advice, thanks!
Creating GPOs for folder redirection PER department? Also allocating a virtual disk for each departments shared drives?
I see how this would be beneficial for backups as well. All Virtual disks would be smaller in size, in comparison to having one large 1TB+ virtual disks.
I would still like to hear more input from anyone wiling to offer advice, thanks!
ASKER CERTIFIED SOLUTION
membership
Create a free account to see this answer
Signing up is free and takes 30 seconds. No credit card required.
I attach storage to my Windows file server VM via the iSCSI initiator inside the guest. It works very well for us. I don't know if that will work with Veeam or not. The advantages that I see are that I can expand the volume on the fly, you don't have the 2 TB limit, and other machines can connect to the same LUN (with proper clustering) or a snapshot of the LUN. You can also easily migrate to a physical server.
You can use Win2008 DFS to distribute the shared folders on different physical volumes.
That would allow to raise the total storage space to ... "no limit" while managing a "classic" storage subsystem (raid 10 arrays to achieve a high IOPS needs or raid 60 arrays to achieve large volumes with not that many random writes)
That would allow to raise the total storage space to ... "no limit" while managing a "classic" storage subsystem (raid 10 arrays to achieve a high IOPS needs or raid 60 arrays to achieve large volumes with not that many random writes)
ASKER
DFS allows you to distribute a namespace across servers, but not physical volumes on a single server. At least this is my understanding, am I correct?
DFS is not usually used to aggregate volumes on a single server...but it can do it.
Basically, it provides a logical view to your physical storage by aggregating "shares" to a single namespace hosted by a virtual server.
In your case, you can define a DFS Win2008 R2 VM to use as many virtual drives as needed and still provide up to 10TB storage.
Basically, it provides a logical view to your physical storage by aggregating "shares" to a single namespace hosted by a virtual server.
In your case, you can define a DFS Win2008 R2 VM to use as many virtual drives as needed and still provide up to 10TB storage.
ASKER
Thank you to everyone who gave input on this thread. We will be restructuring our GPOs and virtual disks to better accommodate backups and logical storage. This is the most beneficial route for our organization.