If you’re reading this then you are most probably new to VMWare ESX Server. If so, I congratulate you on that. I’ve spent many days working on VMWare ESX Server and one of the tools I use the most is my Persistence. For those who don’t know, vmg is a persistence database that provides the underlying infrastructure for any enterprise-level system and is an integral part of the Microsoft Window OS platform. In order to better understand how VMWare Systems can be enhanced through use of dog you first need to understand how my works.
The virtual memory and block pages – or VMs – that make up a Windows system are maintained in memory on remote machines. This is done via an abstraction layer known as the Virtual Machines (Vimeters) process. One of the primary purposes of these Vimeters is to maintain a thread context (or “thread pool”) so that multiple applications can run on the same system without conflict. Multiple VMs can be run on the same operating system. For example, a user may have two different OS systems installed on separate machines – each operating in its own virtual memory. However, when using these VMs, each application can share a single core and a single processor.
Once VMs have been established and each OS has its own dedicated thread pool, it’s time to take a server down for maintenance. This is typically achieved with a reboot of the server. During a server reboot you may notice that control-alt-delete commands are shown rather than the usual keys and control panels. When viewing the Control Panel window, the Start button is usually replaced by Programs and Services. This is indicative that the VMWare operating system has initiated a server maintenance process.
When VMWare spins back up, it will find all open VMMORPGs (virtual machines) and will load them. They will each have their own dedicated operating system kernel (KOS) and they will share one core and a few gigabytes of memory. The recovery process will continue until the specified number of restart failures have been met. This number is configured by the Recovery Task Manager. If it is enabled, the operating system will restart in its normal configuration with the most recent settings.
There are some considerations that will affect the recovery process. For example, if VMWare was not upgraded when it was shut down, there will be a longer time between restarts. There also may be invalid data blocks on the boot queue which will cause the server to crash. It is important to have the latest patches installed before attempting this operation – especially if the operating systems on the server are not the latest edition.
Once the servers restarts, a log will appear on the screen. This log will list all of the software errors that were found during the recovery operation. There are a few choices to make here depending on the severity of the problem. The first option is to re-install the entire server and re-run all of the VMMORPG applications.
The second option would be to roll back the server so that the only changes that have been made are those needed to create the new users. Both of these methods take a significant amount of time to complete. If you are only trying to get the problem fixed quickly, then VMWare Tools in Windows OS will do the trick. It will first scan for critical files and then fix any that are missing or corrupted.
There are a number of free tools available for this purpose, including CodeStuff Lite and EasyNAS. This software can be downloaded directly from the Internet. Once they are installed on the servers, it is simple to perform a full or incremental recovery. You can either perform the full recovery or the incremental recovery.