August 2010 Newsletter
Some of you have told me that when it comes to dealing with upgrades, PTFs, and doing Systems Management you would rather be like Steven Slater, the flight attendant on JetBlue: and just grab two beers, pull the rip-cord, and bolt down the emergency slide to get away from it all. They said he had unruly passengers. Perhaps you have unruly users, have out-of-control performance issues, are unsure what you are really backing up, require PTFs, or are looking at an insurmountable V6R1 upgrade? We can keep you on the plane and away from the slide! This is why you should call iTech Solutions when you have a problem on your IBM i (AS/400, iSeries, i5). We can help you with those annoying upgrades, handling your PTFs by getting you on our quarterly maintenance program, performing a health-check, remote administration, remote hosting, replication, or anything else in the Systems Management or Administration area. We have the experience, the expertise, and the knowledge to help you.
Well we have certainly been busy with new machine installs, i5/OS upgrades, and other projects this summer. Although, I have to say everyone at iTech Solutions has managed to take some vacation time this summer to refresh our batteries. We hope you have as well. August was a huge month for IBM i, with the refresh of the product line with low end Power7 servers. Let me tell you there is nothing low-end about these servers, and I will share some insight into these announcements from the time I spent at Power Systems Champions class in Austin, Texas a few weeks ago.
We have packed a lot of information into this newsletter, and I hope that you find this useful. This issue of our newsletter has five articles. In the first, I want to discuss some things you need to think about for restoring your system. The second article is on the August 17th Power7 announcements. The third article is Hold Job Schedule Entries. The fourth article is a list of upcoming events we will be participating in. The last article is for your reference with updated PTF information.
iTech Solutions can help you improve performance, upgrade i5/OS, perform security audits, implement a High Availability solution, Health Checks, Systems Management, Remote Administration, PTF management, Blade installations, iSCSI Configurations, Backup/Recovery, upgrade an existing machine, or upgrade to a new machine. If you are thinking of LPAR or HMC, then think iTech Solutions. We have the skills to help you get the most out of your System i.
Do you have the right stuff: Saving and Restoring
This article isn’t about John Glenn. It’s about backup and recovery. I am writing this article after doing six recoveries over the last three weeks for clients. Now none of these recoveries were due to losing the primary system. They were all test recoveries. That’s the start of the good news, and probably the end of the good news as well. In two of them, we restored a Save 21 (Full System Save) to their backup machine. Now I will argue this is useless in testing your DR plan, unless you are doing a Save 21 each and every night. You can’t perform a special save of your data for the express purpose of doing a restore with that save. That doesn’t test anything out at all; well perhaps it tests that you are able to put a tape in the drive, enter GO SAVE on a command line, and then select option 21 and hit enter twice. These customers weren’t doing a Save 21 each night, so that means this isn’t a real test. What it was is a check-off on the auditors list of have you tested your recovery. You can tell from my writing that I wasn’t happy doing this, because I feel we are providing false hope in a customer’s ability to restore their system.
When you are really serious about testing your backup strategy, then you have to take your current backups (no matter how out of date or what) and use all the tapes to re-create the system as it was. This is why they call it a T-E-S-T. This is no small feat for some people when I look at how they are doing their backup, but it is a requirement. As an aside, please, if you are unsure, have us come in for a four-hour health check, and we will analyze your backup strategy during that engagement and let you know if you are missing libraries/objects or if the recovery is so complex you are just fooling yourself. This is serious stuff, and as you read this you need to realize that if you are the person responsible for the machine and the backups, failure to properly protect the company could result in your termination (and failure of the company to recover from a disaster).
First, let’s start with how often are you doing a full system backup? I would highly recommend weekly if possible, and monthly if you had to. The least recommendation is quarterly if you have a hard time finding a window large enough to do the backup. Is that window caused by your tape drives being too slow? What tape technology are you using? You need to match the tape technology with the amount of your data on disk, as well as the time constraints of how long the backup takes. LTO is probably the preferred technology in that it is cost effective, fast, and can hold a lot of information.
Some of the things to be aware of if you are on V5R4 and later are you backing up spool files. Will your users require their spool files during a recovery? Are you able to recreate these spool files if they weren’t available?
Do you have logical files in libraries that are over physical files in different libraries, where the library name of the logical file comes before (alphabetically) the library name of the based upon physical file? This is fixed in V6R1, with using the restore deferred (parameter or command).
Are your journals in libraries that get restored after your physical files are restored? If that is the case, then none of your files will be journaled on the restore?
Will you have to have your I_BASE_01 CD available for a restore?
Are you ending the right jobs and subsystems each night so that you are getting all the objects you think you are backing up. When objects are exclusively locked, you won’t get them on your nightly backup.
I had one account that had a customized CL for backing up their libraries. Well, no one has changed that CL for the last few years, and they had a few new programmers join the company as well as changing the version of their ERP software. Guess what happened during the restore? We found out that those programmers libraries and some of the new libraries for their ERP application weren’t getting backed up. A disaster? No, because we tested this out, found the problems, and made the changes. This company is now in a better situation because they did a test and made changes based upon their findings.
These are just some of the questions you need to be thinking about. If you don’t have a system to which you can restore (if you only have one machine), we can bring a machine to your office, or you can just send us the tapes and we can restore it in our office. One of the services that we offer is the ability to test your recovery on one of our machines. We can also host your environment on one of our machines as well if you are looking at doing off-site replication.
If you haven’t done a test this year, you really have no idea the issues that you might have, or the items that may have changed from year to year in how you perform your backup strategy. I urge you to review your backups, contact us if you need help, but most importantly test out your restore.
On August 17th, IBM announced the completion of their Power7 line of servers, with the low end 710, 720, 730, and 740 models, along with the high-powered 795. These models complete the Power Systems lineup of servers with POWER7 chips inside of them. If you remember, the middle of the product set was announced in February of this year, see the iTech Solutions February newsletter.
These new low end machines have an amazing amount of horsepower (CPW). An absolute amazing amount of CPW! I was in Austin Texas at the end of July to learn and play with the machines. I was extremely impressed by their performance, energy savings, small footprint, as well as their cost. The 710 and 730 are small 2U rack mounted machines, but are probably more oriented to the AIX market. I believe most of the IBM i customers will be mostly interested in the 720 and 740 machines. The 720 is 4U high (Rack mounted or desk side) with 1 Socket that can have 4, 6 or 8 cores. A maximum memory of 128 GB (on two riser cards), 4 PCIe card slots and an optional 4 low profile card slots. You can attach external I/O drawers and external disk ONLY on the 6 & 8-core machines. Each core is capable of 10 virtual partitions (or as IBM likes to now call them virtual servers). The electrical power is 100 to 240V. The 740 is 4U high (rack or desk side) with 2 Sockets that can have combined 4, 6, 8, 12, or 16 cores. A maximum memory of 256 GB (on 4 riser cards), 4 PCIe card slots and an optional 4 low profile card slots. You can attach external I/O drawers and external disk on all models of the 740. Each core is capable of 10 virtual partitions. The power is 200 to 240V.
Some items to note:
Ok, so why did I say the 720/740 models are better for IBM i customers than the 710/730 models: you can have more internal storage (disk drives), internal tape drive, twice the amount of memory due to the size of the memory riser cards, and four full-height 4 PCIe slots (710/730 models have no full-height cards).
Remember, there are no upgrades from Power5 to Power7 machines. If you have a Power5 machine, you would have to upgrade that first to a Power6 machine, and then to a power7 machine. So, if you are on a Power5 machine you should contact Glenn to find out what your options are now, before that upgrade path goes away. Also, if you are still on V5R4, your only option is to go to Power 6.
It’s a great time to be an IBM i customer with this announcement. Great new technology that is easy to take advantage of. With all the good, there is always something bad. These machines operate at such amazing speeds, and when you have fast CPUs they generate a lot of heat. Well to move the heat, IBM had to put in huge fans that require a very tight seal to properly move the air past the CPUs (this is what moves the heat away from the CPU). The drawback is that IBM doesn’t allow for concurrent maintenance for the PCIe cards now. If you or the CE were to open the machine up (as you have been able to do in the past), you would cause the machine to immediately lose power. Yes, immediately lose power. This is due to the need for airflow. When the cover is opened, the system shuts off to prevent overheating. What this means for us, is less concurrent maintenance. I give this an “F” in the design, because for a few dollars more IBM could have put a hinge in the cover that would have allowed access to the PCIe slots, but kept the seal over the CPUs for proper airflow. For speed, performance, versatility, energy efficiency, I give the machines an “A+”. What this all means is you had better watch your colors. Orange tabs mean concurrent maintenance. Blue tabs mean maintenance must be performed when the machine is powered down.
For a review of your machine upgrades, or to see exactly how inexpensive moving to this new hardware or even Power6 can be, contact Glenn via email or by phone at 718-898-9805.
I love visiting customers because I learn something new every shop that I go visit. Have you ever wanted to hold every job in the job scheduler? Well, I just found out that there is a command that can do this for you. I used to do a WRKJOBSCDE, and then put each job in the scheduler on hold. I do this when I am going to do an i5/OS upgrade, and we don’t want the jobs to start after the upgrade until we have done conversions or any other post installation activity.
The Hold Job Scheduler Entry command, HLDJOBSCDE, allows you to hold an entry, all entries, or a set of entries in the job schedule. Each job schedule entry contains the information needed to automatically submit a job to be run once or at a regularly scheduled interval. All jobs once held will be held until released. When the job is held and then released it will then pick up at the “next” scheduled interval, and not submit for an interval during which it was held.
Of course, there is a Release Job Scheduler Entry command that does the exact opposite, releasing any job scheduler entry which was held. You do need to be careful when using these commands with the *ALL or generic job names, as you are affecting multiple jobs at once. You should look over the job scheduler before executing any of these commands,
to determine if there are any jobs already on hold or review jobs you might not want to release.
Here is a list of upcoming events that iTech Solutions will be partisipating in or that Pete Massiello will be speaking at:
October 3 to 6 COMMON Fall Conference www.common.org San Antonio, TX Topics:
1) Upgrading to IBM i 6.1 & 7.1
2) HMC, FSP, IBM i: You need to know what these do.
3) Performance tuning tricks
4) Creating Virtual IBM i LPARs on hosted IBM i.
October 20 – Long Island Systems User Group (LISUG), www.lisug.org Woodbury NY
October 21 – North Eastern System Technology Users Group (NESTU), www.nestu.com Fairfield NJ. Topic: Upgrading to IBM i 6.1 & 7.1
November 15 to 17 iSeries Dev Conn www.iseriesdevcon2010.com Las Vegas, NV Topics:
1) Everything you need to know to upgrade IBM i (i5/OS) to IBM i 6.1 or 7.1
2) Building virtual i partitions hosted by IBM i – no additional hardware required!
3) Putting the pieces together: Understanding your options with HMC, FSP, and firmware on the IBM i
4) Tips and tricks to improve system performance and save disk space
5) Everything you need to know to get started with the new Systems Director Navigator Console
April 11 to 13 Northeast User Groups www.neugc.org Framingham, MA
|Release levels and PTFs|
People are always asking me how often they should be performing PTF maintenance, and when is the right time to upgrade their operating system. I updated this article from last month with the current levels of PTFs. Let’s look at PTFs. First, PTFs are Program Temporary Fixes that are created by IBM to fix a problem that has occurred or to possibly prevent a problem from occurring. In addition, some times PTFs add new functionality, security, or improve performance. Therefore, I am always dumbfounded as to why customers do not perform PTF maintenance on their machine at least quarterly. If IBM has come out with a fix for your disk drives, why do you want to wait for your disk drive to fail with that problem, only to be told that there is a fix for that problem, and if you had applied the PTF beforehand, you would have averted the problem. Therefore, I think a quarterly PTF maintenance strategy is a smart move. Many of our customers are on our quarterly PTF maintenance program, and that provides them with the peace of mind of knowing their system is up to date on PTFs. Below is a table of the major group PTFs for the last few releases. This is what we are installing for our customers on iTech Solutions Quarterly Maintenance program.
7.1 6.1 V5R4 V5R3
Cumul. Pack 10096 10215 10117 8267
Grp Hipers 10 69 134 169
DB Group 2 14 27 24
Java Group 3 13 24 23
Print Group – 18 41 20
Backup/Recov. 3 16 34 33
Security Group 3 18 15 7
Blade/IXA/IXS 2 16 14 –
Http 2 14 24 17
If you have an HMC, you should be running V7R7.1 If your HMC is a C03, then it should stay at V7R3.2 SP1.
For your Flexible Service Processor (FSP) that is inside your Power 5 or Power5+ (520, 515, 525, 550, 570), the code level of the FSP should be 01_SF240_382. Power 6 (940x M15, M25, & M50 machines, and 8203-E4A & 8204-E4A) customers should be running EL350_071. For Power6 (MMA, 560, and 570 machines) your FSP should be at EM350_071. If you have a Power6 595 (9119-FMA) then you should be on EH350_071. POWER7 the firmware level is AL710_086 or AM710_086 depending on your model.
If you need help with upgrading your HMC or FSP just give us a call. We will be happy to perform the function for you or assist you in doing it. Contact Pete Massiello.