FluidFS

2 Bronze

Installing Second Equallogic

We currently running VMware ESXi and have a PS5000E in production with firmware 5.2.4 and have purchased a PS6100 to replace it (current firmware unknown). We want to use the 5000 in the future as a replication partner. What is the best method to move all the data over to the new SAN and remove the old SAN and not incur any downtime? And is there anything that has to be done on the VMware side?

Solution (1)

Accepted Solutions
3 Argentum

More than likely the firmware of the new unit will be v6.x.x.

If you're wanting to do asynch replication do the following.

The best way to accomplish the migration without downtime would be to add the new PS6100 to the management group (and storage pool) the PS500E is in and allow the volumes to spread across both members.  Next delete the PS5000E from the group (under the Group - Members tab) and the system will automatically evacuate remaining data to the PS6100.  Once that has finished the PS5000E is ready to be reconfigured (the existing configuration will be wiped out as part of the deletion) in a new group, and set as the replication partner.

The only thing you'd need to do on the VMware side in this scenario would be to add the second group to you iSCSI initiator (on the dynamic discovery tab) so that it's ready if you ever need to connect to a replicated volume.

If you're wanting to do synch rep do the following.

The best way to accomplish the migration without downtime would be to add the new PS6100 to the management group (and storage pool) the PS500E is in and allow the volumes to spread across both members.  Next create a new storage pool and move the PS5000E to it, all remaining data will be evacuated to the PS6100.

View solution in original post

Community Accepted Solution
Replies (9)
3 Argentum

More than likely the firmware of the new unit will be v6.x.x.

If you're wanting to do asynch replication do the following.

The best way to accomplish the migration without downtime would be to add the new PS6100 to the management group (and storage pool) the PS500E is in and allow the volumes to spread across both members.  Next delete the PS5000E from the group (under the Group - Members tab) and the system will automatically evacuate remaining data to the PS6100.  Once that has finished the PS5000E is ready to be reconfigured (the existing configuration will be wiped out as part of the deletion) in a new group, and set as the replication partner.

The only thing you'd need to do on the VMware side in this scenario would be to add the second group to you iSCSI initiator (on the dynamic discovery tab) so that it's ready if you ever need to connect to a replicated volume.

If you're wanting to do synch rep do the following.

The best way to accomplish the migration without downtime would be to add the new PS6100 to the management group (and storage pool) the PS500E is in and allow the volumes to spread across both members.  Next create a new storage pool and move the PS5000E to it, all remaining data will be evacuated to the PS6100.

View solution in original post

Community Accepted Solution
Moderator
Moderator

Since you are planning to replace the PS5000 with the new PS6100, the simplest approach is to add the PS6100 into your existing Group (this will allow for a seamless addition of the array into your ESX environment).  However, note the steps below to minimize the performance impact when adding the PS6100 member, and removing the old one.

First, you want to have a copy of the installation guide for the PS6100 (go to the support site, and download the manual it’s under the Resources> Documentation, look for the install guide).  This doc has the wiring and step by step instructions for configuring the member into an existing group.

So in short:

Wire the array into your existing iSCSI network, and (optionally) management network (if you have setup the dedicated management port on the array).

Connect to the array via terminal or use the Remote setup wizard.

Give the new array a name, and configure the eth0 IP address, and add it to the existing group (you will be asked if this is a new group or existing).  Note the new PS6100 will required unique IP's for each eth interface you configure.

Once the array is added into your production group, it will be "un-intilized" i.e., no raid policy is configured yet.  Now go into the GUI and create a “temp” pool (or create one with the CLI).  The temp pool will be used to move the PS6100 array into while the RAID policy is initialized and expanded (reduces performance hit on existing volumes in the production pool). To do this, select the PS6100 member then configure the raid.  

As you walk through the RAID setup, select the temp pool, and select “wait until the member storage initialization completes” check box.

Once the RAID is completed (depands, could take an hour or a few hours, to see the status use the CLI member select membername show command to see the operation in progress or in th GUI)

You can now simply move the PS6100 into your production pool.  The existing volumes will begin to balance over to the new PS6100 member.

Once the PLB (performance load balancer) has completed balancing the volumes access both arrays, you can now move the PS5000 into the temp pool (the time the PLB takes depends on the amount of data and load on the PS5000 and can take 24hours to a few days).  Removing the PS500 into the temp pool, will vacate all the volumes from the PS5000 to the PS6100.   

Once the PS5000 is moved out of production pool, if you plan on doing remote site replication, remove the PS5000 from the group.

Or if you want to use sync replication (replication to the same group, but the member must be in separate pools), you can reconfigure the PS5000 as needed (move to a new pool, reconfigure the RAID policy, remove it from the group and add it back in later, etc.).

-joe

-Joe

Social Media and Community Professional
#IWork4Dell
Get Support on Twitter - @dellcarespro

Follow me on Twitter: @joesatdell 

2 Bronze

While you describe a procedure I performed a few times with great results, I've recently stumbled across this paragraph in the Configuration Guide:

2.1 About member firmware

Each control module in a group member must be running the same version of the PS Series firmware. Firmware is stored on a compact flash card or a microSD card on each control module.

Dell recommends the following:

• Always run the latest firmware to take advantage of new features and fixes.

• All group members must run the same firmware version. If you are adding a new array to a group, update the group to the latest firmware before adding the new member.

 

You didn't mention (and neither did I do it in the past) that the existing array should be updated to the FW revision the new array has. I have never had a problem introducing a new array with newer firmware into a group, yet the Configuration Guide states it is "recommended".

Why is it recommended, and what kind of problem could this avoid?

Moderator
Moderator

That is an excellent question.

There are a few reasons for this recommendations, and one is that miss-matched firmware in a group is only support as long as it take to upgrade the other members.

One problem is that replication partners should be on the same version of the firmware at each site, however, we do support a partner site being one revision lower (or higher) to cover the times when you are updating one group to a newer FW version (and planning on doing the partner site after then 1st site is completed).   However if you introduce a member into a group that is two versions higher than the lowest partner site, replication will break until this is corrected (update firmware).

Additional reasons include possible degraded performance and functionally issues (for instance the newer FW supports features not supported by the older FW).

Also note that adding a member to an existing group the “major” revision must be the same, so adding a v6.x to a v5.x group will not work.  Typically, you order your array with the firmware pre-installed with one that matches your existing group.

If in fact this new array is v6.x, you can downgrade this model to the firmware version that matches the group you want to install it into.  However you need to open a support case for this.

-joe

-Joe

Social Media and Community Professional
#IWork4Dell
Get Support on Twitter - @dellcarespro

Follow me on Twitter: @joesatdell 

2 Bronze

Thanks, that's very helpful!

It might just be me, but this Information is non-obvious and quite important to have. It should really be in the Configuration Guide.

Anonymous
Not applicable

Hello,

FYI: Replication and firmware compatibility info is in the release notes downloadable with the firmware itself.

Information on member and pool management is in the Admin guide.

Regards,

2 Bronze

Michael my Dell remote installation team states that because the 5000 has a lower firmware they do not recommend putting it in the same pool because they say it might not work. They want me to upgrade the 5000 to 6.0.2 then add it to the pool. Is this correct?

Anonymous
Not applicable

5.2.4 is compatible with 6.0.x.  You will have to upgrade it anyways to stay in a supported configuration.  But you can safely add a 6.0.2 to this group.   There is no harm in adding it to the same pool once you have let the RAIDset finish building and enabled all the network ports on the new member.  Double check that the 5000 interfaces are all enabled as well.

You can also add the 6100 to a new pool and move the volumes individually from the 5000 to the 6100.  That will take awhile but its safe to do as well.  Then you can upgrade the 5000 and remove it from the Group safely.  

If you can upgrade the 5000 now before adding in the 6100 that is always the best option.

Regards,

Anonymous
Not applicable

To clarify, it's OK to add it to the same pool since you are planning on removing the 5000 from the group.  If you plan on keeping it in the group or pool for awhile, then definitely upgrading the firmware is a must.

Top Contributor
Latest Solutions