Dell PowerEdge RAID Controller setup for ScaleIO

Thank you! Pawel Everything worked out

Thank you! Only I wanted to find out in which mode to configure hard drives. build RAID 0, 5, 10 or directly, in which mode should the controller be set to n310 н710? There is no information in the document.

Unfortunately I do not download this file due to lack of access to technical support. Can you download this file for me?

Hi,

Every physical disk should be configured as a standalone RAID0 VD (see "perccli /c0 add vd type=raid0").

Cheers,

Pawel

Thank you! Let me ask one more question. We have three dell r420 servers in each sfp + card for a 10gb network, two servers with a h710 controller and 8 2.5-inch hard drives each 1Tb. 1Tb * 8 = 8Tb and one server with h310 controller and 4 3.5-inch hard drives each 2Tb. 2Te * 4 = 8 Tb. Is it possible to configure these 3 servers with their disk capacities in the shared store ScaleO? do I expect to use esxi6.5U1 or should the hard disks be the same in all three servers?

Thank you! But I still decided to replace the server with 4 disks per server with 8 disks and the H710 controller so that all servers were the same

Hello Pawel! I decided to put four disks in each server, tell me how to increase the number of disks in the future, for example by adding one disk to each server or telling which instruction to read it

Hello Pawel! Help me please. Tell me after the deployment of ScaleIO I can migrate the service virtual machines ScaleIO: Manager1, Manager2. TB1, Gataway from the local VMware Datasore to the shared VFMS Datastore created by ScaleIO to work in the vSphere HA cluster. In case of a single host failure, the ScaleIO virtual machines restarted on the working hosts? And the released disks of local Datastore add to the ScaleIO system. I have three servers with four disks. One disk I allocated to each server for local VMware Datastore for deployment of ScaleIO , and three disks allocated for ScaleIO on each server, now I would like to release these disks and join ScaleIO to expand the volume.

ScaleIO VMware

VMware vSphere 6.5U1, ScaleIO 2.0.1.4

3-node cluster

Thank you Pawel for the answer!!! You helped me a lot!!! Tell me please, then, if one of the servers fails, I can deploy ScaleIO virtual machines (SVM) from the backups on the other two working hosts of the cluster, because if the host fails, ScaleIO (SVM) will not work on this host and it will not restart cluster HA as it is located on the local storage. And on one host I have two ScaleIO (SVM) Manager1 and Gataway. Can I use VMware Data Protection to create backup copies of ScaleIO virtual machines (SVM) and in case of an accident deploy them from backup copies on working hosts? Or, in the event of an accident, there are some recovery rules and you do not need to do this? I do not know how to act when the host fails! Can you advise me! I think correctly about the backup copies of ScaleIO (SVM)?

Hello,

Unfortunately, the ScaleIO Virtual Machines (SVM) need to be on the local storage - you should not move them to a shared storage, even ScaleIO volume, as in such a case you can encounter a chicken and egg problem - unavailability of some components will make all elements to fail and unable to recover.

If you really need to use the local devices SVMs are running on, you can partition them and assign one partition to a datastore where the SVMs will be located and add another partition (vmdk) to SDS, but it's not really convenient and will most likely affect performance - I would recommend you to leave the SVMs where they are now, on their local disks.

Cheers,

Pawel

Hi,

it depends on the type of the SVM - if it's acting only as an MDM/Gateway, you can restore it on another machine, but it most cases they also act as SDS's - in such a case restoring the SVM will give you nothing, since SDS needs access to its disks which participate in ScaleIO cluster. If the host is down, the disks are obviously not accessible, so even if you restart this SDS Virtual Machine on another physical machine you won't be able to use it anyway, so there is no point in doing that.

Please remember that ScaleIO is fully redundant, so even losing a single MDM/SDS doesn't indicate a data loss - in such a case you should either try to restore the physical host or if it's not possible - simply install a new host and add it to the ScaleIO system, so the MDM cluster is protected and there is no degraded capacity - ScaleIO vSphere Plugin is the best way of doing that. You can even try to use the same host as previously, please just remember that if you are going to re-use disks previously managed by some SDS you might want to zero the first few MBs of these disk (with 'dd' or some other tool), so they are not recognized as old devices.

Cheers,

Pawel

Thank you Pawel!  Tell me please! How to act when on a working server only the local storage (Datastore) on which ScaleIO SVM works will fail? Unfortunately this disk does not have a fault tolerance - it's RAID-0. What are my actions in this case? How do I restore the ScaleIO SVM virtual machines after replacing the drive. So I wanted to make backup copies for this case. Tell me please! I'm very worried about the system. Thank you that you very much help with your professional advice!