Linux - Virtualization and CloudThis forum is for the discussion of all topics relating to Linux Virtualization and Linux Cloud platforms. Xen, KVM, OpenVZ, VirtualBox, VMware, Linux-VServer and all other Linux Virtualization platforms are welcome. OpenStack, CloudStack, ownCloud, Cloud Foundry, Eucalyptus, Nimbus, OpenNebula and all other Linux Cloud platforms are welcome. Note that questions relating solely to non-Linux OS's should be asked in the General forum.
Notices
Welcome to LinuxQuestions.org, a friendly and active Linux Community.
You are currently viewing LQ as a guest. By joining our community you will have the ability to post topics, receive our newsletter, use the advanced search, subscribe to threads and access many other special features. Registration is quick, simple and absolutely free. Join our community today!
Note that registered members see fewer ads, and ContentLink is completely disabled once you log in.
If you have any problems with the registration process or your account login, please contact us. If you need to reset your password, click here.
Having a problem logging in? Please visit this page to clear all LQ-related cookies.
Get a virtual cloud desktop with the Linux distro that you want in less than five minutes with Shells! With over 10 pre-installed distros to choose from, the worry-free installation life is here! Whether you are a digital nomad or just looking for flexibility, Shells can put your Linux machine on the device that you want to use.
Exclusive for LQ members, get up to 45% off per month. Click here for more info.
We have purchased a new Network Storage for our department. We are now in the process of migrating the data from Old to new. This includes migration of Datastores to another Storage(Netapp).
Ofcourse we have running VMs in these Datastores. That is the internal disk of VMs reside in these datastores. The datastores are connected to VM-Hosts via NFS.
The name of the remote datastore is Netapp and Netapp1. And internal Datastore is datastore1. Netapp-Datastore's size is 3.25 TB.
And total no. of VMs are around 20. We have NFS connectivity to this storage from VM-Hosts. This can be seen in the attachment.
One the new storage is available over network, we should be starting the process of data migration. Once data migration from one storage to another is complete, we should migrate the VMs.
We would like to know if the following plan will work or needs modifications. (We have a free copy of VSphere 5.5. No Vcenter or any other license. We aint lucky enough to purchase them )
1. Migrate (copy) the data from 1 storage to another. This includes the datastores as well.
2. Once the data migration is complete, then we bring down all the clients as well as VMs and their respective VM-Hosts. Before bringing down the VM-Hosts, we will have to add the new Datastores under Storage using VSphere. Question1: Should the existing storages be deleted at this stage? We think yes, because we have to add the new datastores with exact same name as the old datastores. In this case Netapp and Netapp1. Only then the VMs will be able to start. As for a particular VM, the VM-Disk is accessed via Datastore name. E.g. In the second attachment, we can see the VM-Properties show, the internal disk is represented as [Netapp]<Name of the host>/Name of the Folder. So we delete the 2 Datastores entries and add them again with new IPAdress but the same old Name Netapp and Netapp1.
3. Bring down the ESXi Hosts.
4. Re-start the Hosts. This will now include only the new Datastores. Make sure that these are now accessible.
5. Start the VMs Question2: Will these start without any adjustment or modifications? Or we should be doing something extra for it to work? Will VM detect that the hosts are now accessible from the datastores running on a new Storage?
Though we are using a free version of VSphere, if the above method does have limitations or is absolutely not possible then we will certainly ask management to consider buying VMWare licenses.
Kindly share your expert comments or any modifications. Your comments will be highly appreciated.
Hello Experts,
We have purchased a new Network Storage for our department. We are now in the process of migrating the data from Old to new. This includes migration of Datastores to another Storage(Netapp). Ofcourse we have running VMs in these Datastores. That is the internal disk of VMs reside in these datastores. The datastores are connected to VM-Hosts via NFS.
The name of the remote datastore is Netapp and Netapp1. And internal Datastore is datastore1. Netapp-Datastore's size is 3.25 TB. And total no. of VMs are around 20. We have NFS connectivity to this storage from VM-Hosts. This can be seen in the attachment.
One the new storage is available over network, we should be starting the process of data migration. Once data migration from one storage to another is complete, we should migrate the VMs. We would like to know if the following plan will work or needs modifications. (We have a free copy of VSphere 5.5. No Vcenter or any other license. We aint lucky enough to purchase them )
1. Migrate (copy) the data from 1 storage to another. This includes the datastores as well.
2. Once the data migration is complete, then we bring down all the clients as well as VMs and their respective VM-Hosts. Before bringing down the VM-Hosts, we will have to add the new Datastores under Storage using VSphere. Question1: Should the existing storages be deleted at this stage? We think yes, because we have to add the new datastores with exact same name as the old datastores. In this case Netapp and Netapp1. Only then the VMs will be able to start. As for a particular VM, the VM-Disk is accessed via Datastore name. E.g. In the second attachment, we can see the VM-Properties show, the internal disk is represented as [Netapp]<Name of the host>/Name of the Folder. So we delete the 2 Datastores entries and add them again with new IPAdress but the same old Name Netapp and Netapp1.
3. Bring down the ESXi Hosts.
4. Re-start the Hosts. This will now include only the new Datastores. Make sure that these are now accessible.
5. Start the VMs Question2: Will these start without any adjustment or modifications? Or we should be doing something extra for it to work? Will VM detect that the hosts are now accessible from the datastores running on a new Storage?
Though we are using a free version of VSphere, if the above method does have limitations or is absolutely not possible then we will certainly ask management to consider buying VMWare licenses.
Kindly share your expert comments or any modifications. Your comments will be highly appreciated.
You've got 20 VM's running on what presumably is decent hardware, with similarly sized network storage. And you can afford to run both of these systems in parallel, while you do a migration. You say you "aint lucky enough" to have licenses?? Luck has nothing to do with it...you need to buy the licenses and pay for support. You're in a decent sized environment, so paying what boils down to a small amount for support/updates is the first thing I'd do. The second thing I'd do would be to contact VMWare support, and get them to tell you what a good procedure is, and follow it. You're obviously running some sort of business with this hardware/software, and don't tell us what versions/distros of Linux the guests are running....so paying for support is worth FAR more than having what could be a major outage for your business.
If you don't want to follow that advice, then do some testing (as you should anyway). Bring up a new VM and run whatever-service-you-want on it. Then follow your procedure to see if it works and what changes (if any) need to be made. Because if it works for ONE machine, it'll work for the other 20. Because until you test and KNOW what works, everything is speculation. Take backups first, and make very, VERY sure they all work. Document and come up with a roll-back plan.
I certainly agree to your point of upgrade and having better licensed software. And maybe that is the reason I am gathering the relevant information by explaining the scenario.*
I am looking for strong grounds so that I will be in a position to convience management (*this time*).*
Wenn it is technically not possible or maybe not has loopholes then I have a strong ground to push things and that is what actually I am looking for. If the above method is technically not possible without better software version then thats all what I want to know.
I respect and agree to your point of view but be kind and comment in terms of the above procedure. As far as testing is concerned, once the storage is setup, there will be a test phase for sure. And hence I wish to know if the above procedure is OK. Or if there is other better method. I have previously cloned the VMs from one ESXi to another having same storage. but that time I only have copied the 2 .vmdk and .vmx files and I was able to start VM on a different ESXi-Host. But this scenario is somewhat different. As in this time whole storage is going to be migrated onto different hardware. In that case i think I need not have to recreate VM-Structure by (Add to inventory-step). I just start the VMs as usual and expect them to run. That's the question.
Hello,
I certainly agree to your point of upgrade and having better licensed software. And maybe that is the reason I am gathering the relevant information by explaining the scenario.* I am looking for strong grounds so that I will be in a position to convience management (*this time*).*
Running a business on commercial, unlicensed/unsupported software is never a good thing. The way you explain it is, "If things break right now, and I can't figure it out, it'll be down until I piece everything together, ask on forums, etc., etc. Or, we can get support and I can make a phone call and get the people who wrote this software and support it all the time working for us."
Quote:
Wenn it is technically not possible or maybe not has loopholes then I have a strong ground to push things and that is what actually I am looking for. If the above method is technically not possible without better software version then thats all what I want to know.
I respect and agree to your point of view but be kind and comment in terms of the above procedure. As far as testing is concerned, once the storage is setup, there will be a test phase for sure. And hence I wish to know if the above procedure is OK. Or if there is other better method. I have previously cloned the VMs from one ESXi to another having same storage. but that time I only have copied the 2 .vmdk and .vmx files and I was able to start VM on a different ESXi-Host. But this scenario is somewhat different. As in this time whole storage is going to be migrated onto different hardware. In that case i think I need not have to recreate VM-Structure by (Add to inventory-step). I just start the VMs as usual and expect them to run. That's the question.
I understand the question, as I did previously; did you understand my first response??
Again: there is speculation and "This should work", and actually doing it to find out. So when you get the new storage, go through your process and find out. Yes, in theory it SHOULD work, but you have still not given any details about things....not the storage used, version of VMware, Vsphere, what the virtual servers are running (Linux? Windows?), etc. We can't guess.
Run a test and find out. If a test is all you're looking to do, there's no reason at all why you can't perform such a test now. Add in more network storage from whatever source you'd like for a test...it supports iSCSI and NFS as options, and you can set those things up on pretty much any flavor of Linux. Do your migration/testing to that...speed isn't going to matter for a test.
So basically you can migrate the boxes one by one, because the script will make a snapshot of the box and the destination could be the new storage that you already setup.
It is a painful process to be frank but usable with relatively short downtime.
Most certainly I would not do the whole lot in the same time.
Also if your data is not changing then this will be the best solution.
If the data is changing on the hosts like you got a local mysql or MongoDB or whatever then I'd suggest to shut down the box, create a backup of the machine with this solution and then start the box on the new storage.
ESXi will complain of the box because at the start-up because it thinks that it is a new box that was either moved or copied, here you could choose that you moved the box not copied otherwise ESXi will change the network card MAC address and then E.G: it is an old CentOS, then you would need to swap the network card MAC under the udev, but this won't be a trouble under Debian.
However I have done this several times to move boxes like this without the venter license.
The first few boxes are a pain but after when you know how to deal with this then it is relatively easy.
Thanks I shall surely try this out.*On the otherside, i wish to as well present VMWare licensing options to the management. As said earlier we have VMWare Vsphere 5.5 Version installed on ESXi-Hardware.What will be the required licenses to purchase? Is VCenter a must? Is the upgrade to newest version prior to purchase licenses a must?*Please suggest.
Thanks I shall surely try this out.*On the otherside, i wish to as well present VMWare licensing options to the management. As said earlier we have VMWare Vsphere 5.5 Version installed on ESXi-Hardware.What will be the required licenses to purchase? Is VCenter a must? Is the upgrade to newest version prior to purchase licenses a must?*Please suggest.
I will, AGAIN, suggest you contact VMWare and ask them about their licensing terms.
Thanks I shall surely try this out.*On the otherside, i wish to as well present VMWare licensing options to the management. As said earlier we have VMWare Vsphere 5.5 Version installed on ESXi-Hardware.What will be the required licenses to purchase? Is VCenter a must? Is the upgrade to newest version prior to purchase licenses a must?*Please suggest.
The vCenter is a must to do live or cold storage migration.
The basic 3 node type version is called essential, cannot do live storage migration, only the normal version could do this.
With the essential version you need to shut down the vmachine and migrate the server to the new storage and then start the virtual machine there on the new storage.
To be frank this is very basic to compare the price, but at least this is a cluster solution and works, but storage wise it is just basic license.
With the solution that I mentioned it is almost the same, but painfully complicated because of the raw edit part via the ssh on the ESXi host itself.
LinuxQuestions.org is looking for people interested in writing
Editorials, Articles, Reviews, and more. If you'd like to contribute
content, let us know.