Factorytalk activation when there is a RAID 1 in the PC.

JesperMP

Lifetime Supporting Member + Moderator
Join Date
Feb 2003
Location
ᚴᚬᛒᛅᚾᚼᚬᚠᚾ
Posts
16,323
Hi.

This is one of these questions that would have been nice to ask on a Rockwell forum if there existed one.

This is the scenario:
I have RSLinx OEM on a customer PC, and I have activated it via the web. I chose to let the activation be linked to the HD. So far so good.

Later the customer informs me that there is actually a RAID 1 in the PC = 2 HDs in mirrored configuration. This I hadn't even noticed.
The customer reasonably expects the RAID 1 to help him get a more reliable system.

When I used the tool for activation, there was only one "Host ID" available for the hard disk(s).
This makes me speculate about where this Host ID comes from, and the possible implications.

1. The Host ID is linked to the RAID controller.
This means that if a HD fails, the Host ID is still valid and the application will run.

2. The Host ID is linked to one of the HDs.
This means that there are 50/50% chance that if one of the HDs fail, the license is still valid.

3. The Host ID is derived from both HDs to some kind of formula.
This means that if either of the HDs fail, the license will be invalid, and the customer is screwed.

It would be nice to know which of the scenarios is the right one.
The online help in Factory talk activation doesn't mention RAID disks.

Btw, if I had known about the 2 HD's I would have chosen the NIC as Host ID.
 
My guess would be 1. The little bit that I have worked with RAID tells me that all the disk are the same. It doesn't matter if you have 2 disk or 4 disk they are all the same.
 
Screw the hd version and key it off the network card...they changing that a lot are they??

And theres always calling AB on the techsupport line for those answers.
 
As production is running 24/7 now, it does not look like that I will get a "window" to mess with the PC in the reamining time that I am here.
I just would like to know what the truth is. They could put it in the online help or something, under "things you should know".
 
Truth?? the computer I worked with that had RAID drives was a pain in the rump....And that wasnt with Rockwell software... It doesnt like dual boot computers so why should it like 2 primary hd's??


It would appear its keyed to the raid controller card not the drives so it may work... The old style activations AB used to use would have gotten corrupt... but the current text file is more forgiving.
 
Magic 8 Ball said:
The old style activations AB used to use would have gotten corrupt
I have Siemens licenses on the same PC, which uses the same copy protection scheme as ABs old EVRSI. And it seems to work with no problems.
No offense, but I really like to KNOW, and not to GUESS.

Magic 8 Ball said:
It would appear its keyed to the raid controller card not the drives
Can you elaborate a bit on that ?

edit: And the same customers has for years run with Rockwell software with the EVRSI copy protection on other PCs with RAID 1.
The point is that this is their (and mine) first go with FTA copy protection on a system with RAID 1.
Is there a potential problem or not ?
 
Last edited:
Jesper,
Great question - I'm curious to know how this works as well. My real answer is, "I don't know". I was hoping one of the Kens would come to the rescue here.

This really depends on how strict/lenient AB chose to be with what ID they used in the generating hash function. This is not something that they'll be likely to want to document thoroghly for exactly the reason that it exists. The "building blocks" that they have to work with aren't too different for a "normal" or RAID setup. You have: controllers, channels, volumes (logical), and disks. If they were going to be really mean they would throw everything they can read into the hash - controllers, channels, disks, etc. This is an unlikely choice as it would cause undue support/user frustration. A change to pretty much anything related, RAID or not, would hose you. More likely they either use the controller id (suggestion #1) or a logical volume id (like suggestion 1 in terms of fault tolerance), which shouldn't change for a drive rebuild, but could for, say, repartitioning your (single) hard drive. It would likely take more work for them to implement #2 or #3. I don't want to say that they're not possible, though, because it depends on how "deep", or how much info, the "get hardware" function returns. IMO #2 is unlikely. Any small change in input to that hash function will yeild a very different result by design.

btw, low level tech support would only know how this works if the approach is so strict that it causes constant problems. The dev team is who to ask (fingers crossed).

JesperMP said:
Hi.

This is one of these questions that would have been nice to ask on a Rockwell forum if there existed one.

This is the scenario:
I have RSLinx OEM on a customer PC, and I have activated it via the web. I chose to let the activation be linked to the HD. So far so good.

Later the customer informs me that there is actually a RAID 1 in the PC = 2 HDs in mirrored configuration. This I hadn't even noticed.
The customer reasonably expects the RAID 1 to help him get a more reliable system.

When I used the tool for activation, there was only one "Host ID" available for the hard disk(s).
This makes me speculate about where this Host ID comes from, and the possible implications.

1. The Host ID is linked to the RAID controller.
This means that if a HD fails, the Host ID is still valid and the application will run.

2. The Host ID is linked to one of the HDs.
This means that there are 50/50% chance that if one of the HDs fail, the license is still valid.

3. The Host ID is derived from both HDs to some kind of formula.
This means that if either of the HDs fail, the license will be invalid, and the customer is screwed.

It would be nice to know which of the scenarios is the right one.
The online help in Factory talk activation doesn't mention RAID disks.

Btw, if I had known about the 2 HD's I would have chosen the NIC as Host ID.
 
Last edited:
Great question, and I'm curious now. Next week I am looking into a server for several reasons. I would not dream of a server without RAID. So I am very inderested in your situation.
 
The other issue you have has to do with how is the RAID being implemented. Software RAID, cheap hardware RAID, real hardware RAID.

Many computers these days offer RAID support on the motherboard. This is basically only a slight step above "software" raid. It is possible that ID in that situation could be coming from the RAID chip or the motherboard.

"Real" RAID uses a separate controller card.

While I don't know the answer I would guess that this might make a difference.

OG
 
Good points. I typically distinguish between "cheap" and "real" RAID based on dedicated controller memory and XOR processor. Separate controller card doesn't always mean better, but it's a good rule of thumb.

I doubt this has much impact on the post, but it's very possible that it could.

Operaghost said:
The other issue you have has to do with how is the RAID being implemented. Software RAID, cheap hardware RAID, real hardware RAID.

Many computers these days offer RAID support on the motherboard. This is basically only a slight step above "software" raid. It is possible that ID in that situation could be coming from the RAID chip or the motherboard.

"Real" RAID uses a separate controller card.

While I don't know the answer I would guess that this might make a difference.

OG
 
I've used both versions of the hardware type. I much prefer real RAID, with hot swappable drives. More expensive, but more reliable.

One customer asked why the box was so expensive and I pointed to the drives and explained how it worked. While doing so I pulled a drive and handed it to him. Replaced the drive, and pulled the second. The whole time he was running his database. He was happy to pay for the server after that. Then he said, "Does this mean that I can quit doing backups?"

Insurance agency, nope.
 
I have investigated this FTA + RAID stuff a little further.
It turns out that the Host ID is not related directly to the disk itself, but to the volume serial number of the C: partition.
On the PC there was both a C: and a D: partition, but only the serial no of the C: partition could be selected as host ID.

So I assume (verification pending) that either of the HDs can fail, and the activation will still be valid.

One thing is that I thought that the serial number was related to the HD itself, so one could repartition the HD without the license being lost. Now I am not so sure.
 
Have you contacted Rockwell Technical support? They should be able to answer this. If not, then they should be able to do some testing to resolve.
 
I had the same question to ask Rockwell,

We upgraded to a raid server, RSveiw32, power32, linx prof + ADS
XP Prof.
Rockwell told me there was no problem & I installed everything including 7 floating active display veiwer licences, this system has been in for about 3 years now & had no problems, never had a disk go down so don't know what would happen but rockwell were confident there would be no problems.
 

Similar Topics

Hi Folks, who know why after start FactoryTalk activation manager V4.03, click "Get New Activations", it displays Is not I appreciate with any...
Replies
0
Views
63
Rockwell Tech Support seemed to have hit a wall with this question. Already updated the version to 5.00.13 per their suggestio but am still...
Replies
2
Views
225
Hello all, I have the misfortune of needing to support some old firmware version 1756 controllogix machines in rslogix 5000, as well as some...
Replies
16
Views
965
I have a Win10 host that crashed this weekend. I reinstalled windows and Factorytalk Activation. Now my VM won't pull the licenses. Host and VM...
Replies
4
Views
866
Can someone provide me activation for FTV studio and client for V12 please
Replies
1
Views
1,008
Back
Top Bottom