Shulou(Shulou.com)06/03 Report--

In the actual enterprise IT environment, there is no need to update and replace the hardware, and sometimes the capacity needs to be expanded when the storage is full. For the applications on WSFC, we can use storage migration technology to replace disks in Hyper-V itself, but for other cluster applications that do not have migration technology, what should we do?

Today, let's take a look at the expansion and replacement of the data disk in the cluster. this time we assume that a file server service based on the cluster has been used very well, but the data disk has not been planned at that time. Full, how to replace the disk while keeping the original data

When faced with this problem, Lao Wang first had two ideas in his head to skip.

Copy the existing contents of the file server to the new disk, dismantle the file server cluster role, and rebuild the cluster role based on the new disk.

Add disks for cluster roles, migrate file server content, modify dependencies, and replace disks

The first method, maybe, involves some permissions to rebuild the file server role, and it is not very professional, so it is generally not recommended

The second method is feasible and convenient, with less downtime.

In addition to the two methods that Lao Wang thought of, there is another way for the cluster. By repairing the cluster disk, you can directly add and replace the cluster role's disk. In the 2003 era, this can be done through Cluster Server Recovery Tool. It is also convenient to use the built-in cluster disk repair after WSFC 2008.

Recently read a foreign blog to know this method, considering that there may be many such needs in China, so sort it out and share it with you.

We will do experimental verification on the second new replacement and the third repair and replacement respectively. This verification uses the 08R2 cluster environment, and the subsequent 1216 does not change this function too much.

Environment introduction

DNS&iscsi

Lan:10.0.0.2 255.0.0.0

Iscsi:30.0.0.2 255.0.0.0

08node1

MGMET:10.0.0.3 255.0.0.0 DNS 10.0.0.2

ISCSI:30.0.0.3 255.0.0.0

CLUS:18.0.0.3 255.0.0.0

08node2

MGMET:10.0.0.4 255.0.0.0 DNS 10.0.0.2

ISCSI:30.0.0.4 255.0.0.0

CLUS:18.0.0.4 255.0.0.0

Currently the cluster is running a devtestfs file server role, using cluster disk 3, with existing software shares

The cluster disk letter is E and there is only 100m of space left.

The idea of the new replacement operation flow is as follows

Storage device allocates new storage LUN to cluster nodes

The cluster node identifies storage disk, online, initialization, partition, and gives a random drive letter that is not E

Use robocopy,xcopy, Explorer, permcopy,fsmt and other tools to copy and copy the contents of the existing file server cluster disk. If you need copy permission, you need to pay attention to it here.

Offline file server role, where downtime begins

Add disks to cluster available disks

Add cluster available disks into file server available storage resources

Modify the file server dependency to point to the new disk

Delete the old disk from the file server, and delete the old disk from the cluster available disk

Change the new drive letter to E

Online file server, end of downtime

Here, according to Lao Wang's attempt, if the file server replacement does not change back to the original drive letter, there will be a problem with sharing after replacement, and you may need to reconfigure the sharing. If you think reconfiguration doesn't matter, you can omit step 9. If you are afraid of trouble, if you want to succeed, it is recommended to do it all at once and modify it back to the original drive letter.

The time node comes to the second step

Each node gets the newly allocated storage and has completed the partition formatting

Copy the existing data to the new disk. Here, in the actual enterprise environment, you can issue a maintenance notice in advance to prevent the copy from generating new data.

Here, Lao Wang does not consider the issue of permissions, but directly copy it by the resource manager. If you need to migrate with file permissions, you can check the blog that Lao Wang has written before.

Offline file server resources for new disk replacement. This part can be completed quickly if the thinking is clear, and the downtime will not be very long.

Add a new disk to the cluster available data disk

Add cluster free disks into file server available storage resources so that file cluster server roles can modify dependencies to new disks

After adding add storage, a dialog box opens to select the cluster available disks we added

There are two data disks available in the current file server cluster role. In fact, if replacement is not considered, it is simply to expand the capacity of the existing file server. We will directly add a disk to the existing file server, and then control the uploading of new data to the new disk. We can also use the same access path.

Modify the file server cluster resource dependency to point to the new cluster disk

Delete old disks from the file server

Cluster free disk delete old disk

Modify the file server new disk drive letter to E

Online file server role, end of downtime

The file sharing server can now be accessed normally, but the data disk has been replaced with the upgraded new disk

Through this upgrade method, we can successfully complete the expansion and upgrade of the file server data disk.

The slightly more complicated one is to understand the process of disk replacement in the cluster.

It should be noted that before starting to deal with the disk replacement operation, Lao Wang suggested that you must take the file server role offline, which is the safest way to do so, because the contents of your files have been copied, and the downtime for replacement here is very short. Don't be reluctant to give up. The actual test replacement operation completes most smoothly when the cluster role is offline.

New-content migration-cluster disk-file server available storage-modify dependency-delete old cluster disk-online cluster role

Next, let's take a look at the third way of repair and replacement, which operates as follows.

Storage allocates new storage to cluster nodes

Cluster nodes go online, initialize new storage, format partitions, and assign a drive letter other than the existing E.

Use robocopy,xcopy, Explorer, permcopy,fsmt and other tools to copy and copy the contents of the existing file server cluster disk. If you need copy permission, you need to pay attention to it here.

Offline existing file server role, downtime begins

Click repair disk and select a new disk

Online File Server role

The time node comes to the second step

Each node gets the newly allocated storage and has completed the partition formatting

Copy the existing data to the new disk. Here, in the actual enterprise environment, you can issue a maintenance notice in advance to prevent the copy from generating new data.

Here, Lao Wang does not consider the issue of permissions, but directly copy it by the resource manager. If you need to migrate with file permissions, you can check the blog that Lao Wang has written before.

After the copy is completed, the existing file server disk is offline and the downtime begins.

Right-click on cluster disk 3 for more actions-repair

Select the new disk that copies the completed data

After the repair is complete, the new disk becomes clustered disk 3

Online cluster disk 3, you can see that the drive letter has automatically changed to E, so this repair wizard can automatically map the disk information of the cluster role that was in use to the new disk.

Online file server role, found that all shares have been restored normally

Through the repair and replacement wizard, it automatically helps us to remove the old disks from the file server cluster role and from the cluster available disks

Through the above two experiments, I believe you can understand how the file server cluster role should migrate disks.

We can find that although the two methods can achieve the same effect

But there is no doubt that the second method is simpler and easier to understand.

It is equivalent to saving us four steps to replace the cluster disk by using a wizard to repair the disk.

As long as you say, to replace or expand the disk, directly add a new disk, copy the content, click repair, come on, done

It's so simple and convenient.

For file server cluster role migration, if your array supports direct expansion, it is best to operate directly at the storage level and expand the cluster disk offline. You can also choose the second and third method introduced by Lao Wang today. the second method may be more convenient for you to understand the process of replacing and updating disks by cluster roles, and the third method is more simple and convenient. I hope it can bring harvest to the interested friends.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.