[MUSIC] As shown here, this is a hell of a look at IPO face. First, the system allows the nucleus, then it locates the master catalog and the other data sets. z/OS is then initialized by using SYS1.PARMLIBparameters. Then z/OS creates the other system adverse spaces. z/OS IPL is now complete but this is simply the big picture. Next, I want to explain the IPL process with a little bit more detail. This page is showing you the four steps and what happens when your IPL system. It is a little bit more complicated than that. I will attempt to try to explain the IPL process in a way that can be better understood. An IPL initial program load is the act of loading a copy of the operating system from disk. That is on your disk drives that you specify with the load address we mentioned earlier. It is putting your CPU central storage and it's executed there. Before an IPL can take place, your system must be powered on and a power reset must have been completed successful. The POR does stuff like set up the LPAR, set up storage boundaries, loads lCDS, loads code in the channels, loads code in the PUs, and so much more than we can discuss here. Once the POR completes, then the LPARS are activated. With the LP activation, this step gives each LPARS all the storage and PUs assigned to them. Then finally the IPL or initialization process starts, we're going out to a console and then pointing to the SYS device specified by the load address. Your system will go out to that DASD address which contains the bootstrap record and reads record for on the DASD device. Record for is the IPL control program known as IEA IPL00, which is simply the IPL text. It is load into memory at this time. This step is called the hardware phase of IPL and it's the first step that we do. Know when your system program of formats the DASD device to be what they're providing is the IPL text or record fou. No other z/OS data set will have this information on it. z/OS is not set up so your system has very little intelligence. So the IPL is always based on the load address that you tell it. If you can't get to the IPL device, your IPL will fill with a disabled waste A code. But if it's successful, the IA IPL000 then invokes a series of program called almons, which we mentioned earlier, the IPL resource initialization manager. It comes from the cyst of that one nucleus. This is called the software phase. It then begins by transferring control to the almons. IPL resource intialization manager. The nucleus is now loaded and the master JCL of this member is used. The almons creates the common area in central storage. The creation of the common system error is done by locating the master catalog and the other system data sets and then initializing z/OS using the information contained in it. Since that one part of life is the next step to build the first address space and create the other system address spaces. The first address space in z/OS will always be the master address space located at address 001. The LDF is then read which contains IO configuration for the LPAR that you just IPL-ed. For each device, a UCB is created in SQA, the system Q erro. The UCP is mapped to a UCW and z/OS is now ready and available to use this device. The last part of irooms is to load the nucleus initialization program into real memory. NIP establishes the nucleus and builds system address spaces needed later by z/OS. It also invokes programs to do resource initialization are simply known as RIMs, the resource initialization program. The nucleus stays in real storage and it's non-pageable. The boot step area and low storage will be overlaid by the prefix save area or what we call the PSA. PSA has an important role in interrupt handling. Since z/OS is not up yet, any errors will be presented to the operator with a disabled waste-A-code. Prior to having an HMC console, this could be very frustrating. In the past, you had to have a NIP console to reply for the NIP messages. But the problem was that message only stayed in the console for a short period of time. After that it would time out and go to another console, then it would time out again and go to another console that you had specified. But the message will remain on the previous console and if you miss that time period it could be very, very frustrating like I said. So as you can see this could be very confusing. Finally, if no console was found, you were given a double 07 disable waste-A-code. Today, if z/OS cannot find a console, it simply goes to HMC and it stays there. It does not go to another console.Thus you no longer see the 007 disable waste-A-codes and no longer do you have to worry about missing the NIP messages. During the NIP stage, the system might prompt the operator to provide system parameters that control the operation NVS. The system also issues information messages from other operators about the stages of the initialization process. NIP is complete when you see the message specify system parameters, but no, all replies in NIP must be responded with R the message number. And then what the message is going to be. And that's going to be done to the command line since z/OS does not activate at this point. Most customs today use automation so most likely you will not see any prompts. The next thing that you should see is up and running z/OS system. Congratulations, you are now an expert on z/OS IPO. That was a lot, but I want to show you that a lot happens in the IPO. No need to try to memorize this. So this is a wrap for now. In the next session, we'll continue our discussion about the load excec and the load PAR members. [MUSIC]
