Jump to content


Photo

yosemite 9.1 DR ISO fail

9.1 disaster recovery iso network interface

  • Please log in to reply
3 replies to this topic

#1 Bereznai Tibor

Bereznai Tibor
  • Members
  • 3 posts

Posted 31 May 2012 - 02:53 AM

Hi,

I am running yosemite 9.1 on a windows network. All servers has two interfaces - one for data and one dedicated for backup. Running DR on win 2008 server previously with yosemite version 9.01, using network (not local) recovey, was OK - the server runing the recovery process on, found master server after ISO image was loaded. The situation changed after upgrading to 9.1 I know, that network interface binding policy changed, so after loading ISO image, when testing interface, the master server can not be found. Meanwhile the master server can see the recovery server, even the recovery server can be pinged on the backup interface, but the recovery process fails. At this stage, the recovery server`s properties can be displayed on the master server, and the connections tab says, that only the backup interface is bound, but the data interface is not. Checking the address tab, there is the explanation - the backup lan interface is OK, but the IP address of the data lan interface of the recovery server is 1.0.0.1 or 1.0.0.2 - what is obviously wrong.
So the question is: how to manage network based DR recovery in a multihomed server environment ?
Should I downgrade to 9.01 ?

Thanks in advance.

Tibor

#2 Ben Kloos

Ben Kloos
  • Barracuda Team Members
  • 173 posts

Posted 31 May 2012 - 10:53 AM

If I understand correctly, when you are in DR recovery mode the machine you are recovering does not have an appropriate IP assigned and therefore can not communicate with the master server. By the way you may have noticed but in DR recovery mode you can press shift+F10 to get to a command interface so you can run ipconfig ping etc.

#3 Bereznai Tibor

Bereznai Tibor
  • Members
  • 3 posts

Posted 04 June 2012 - 10:01 AM

Hi Ben,

Thank you for your reply. I tried your advice, and I am facing new problems. I was trying to recover a win2008r2 again, but on different hardware than previously. Previously, as mentioned before, the problem was the IP address on one of the LAN adapters, but that was an old Intel S5000 PAL server. On the new Intel S5500WB with dual integrated lan and an entry level hardware raid module called AXX4SASMOD, even the adapters were missing at all. No lan, no disk.. The recover process was performed on the same hardware (S5500WB) than the ISO image was generated on. I tried to load the appropiate drivers - the lan and the raid driver as well, it said: driver loaded successfully - but nothing.happened. I can't see any lan (shift F10 -> ipconfig /renew ->"no adapter is in the state permissible for this operation") and no hard drives can be seen, so restore can not be completed - not even using local media.
I really don't know what to do - is there any contraindication of downgrading to 9.01 ?

Regards

Tibor

#4 Ben Kloos

Ben Kloos
  • Barracuda Team Members
  • 173 posts

Posted 04 June 2012 - 12:52 PM

Well the way it should work is that recovering to similar hardware should work because the drivers that were collected at the time of boot image creation should all work with the recovery hardware. One gotcha is that you must have created your logical volumes in the raid controller config before starting the DR process but you probably did do that.

There is a problem we have been tracking with the Microsoft PE utility which is worth mentioning. PE is the mini-OS which we use to run our DR in Vista, 2008, 2008R2 and Seven. This min OS loads some generic drivers to get up and running and we load some additional ones that were collected when the boot image was created. In some cases we have noticed that the drives that PE loads overwrites the drives that were collected at time of boot image creation which can cause some undesirable results. If you have tried explicitly loading the driver for your RAID controller and still can't see your volume(s) I can only assume you are running in to this issue. Fortunately you can recover any system to the same degree that DR does even if DR fails. The general procedure is to reload the same OS using the same machine name, patch the OS to the same level it was at time of backup, install YSB, import your most recent backup and perform a full restore. This will bring the system back to the state it was in at the time of backup.