The end of the rainbow

Somewhere over the rainbow, bluebirds fly…

Yes! The JS12 is still rockin and rolling so we are finally out of the woods and can once again rely on the trusty i to keep going and going.  There are still a couple of mysteries to solve: The bang! light on the SAS Connectivity Module when the tape drive is plugged in (it works though) and the mysterious “Message ID . . . . . . :   CPF9E7F” i5/OS usage limit exceeded message.  — Current processor usage across all active i5/OS partitions is 2.00  — Current processor usage for the local partition is 1.00.

Since I only have a single processor allocated to the i partition and and only one i partition configured, IBM and myself haven’t been able to figure out why i5/OS is still reporting 2 processors active for i5/OS.  PTF to come, no doubt.

Looking into some Ruby 1.9 issues and some other odds and ends while still trying to make ends meet by consulting.  At least I have a machine I can work with and I can, once again, pull the plug on the 270.

Hope to report on some CRUD frameworks in an upcoming post.  Busy as ever, but more productive now that I am not burning up time trying to keep the JS12 running…

Posted in Uncategorized | Tagged , , , | Comments Off on The end of the rainbow

Out of the woods

The blade has now been running 5 days without a hitch so I am thinking that I have finally gotten a stable JS12.  The i partition has been running successfully and I just installed SLES 11 last night.  I am still working through some issues with *that*: It appears to have some issues with graphics so I haven’t been able to start VNC successfully yet.  But, by and large, the blade appears to be rockin and rollin now.

I want to play a bot more now.  I need to get back to my JRuby projects and get a JRuby application running on this Power 6 box and, with the Linux partition available, I’d like to try a PPC64 compile of Asterisk.  If I can consolidate more servers, I’d be able to justify having this big space heater in my basement.

More to come.  Stay tuned!

Posted in Uncategorized | Tagged , , , | Comments Off on Out of the woods

It’s….it’s….ALIVE!

Spent the morning installing fixpack 21 for VIOS.  It went without a hitch.  I started the blade and then once it was up and running, allocated to tape drive to it and started the partition.  The media tray is IN so just about everything that could force this thing to drop is hooked up at the moment but it IS running.

My only concern at the moment is a bang! light on the SAS Connectivity Module (external port) that might indicate an issue with the tape drive or cable (since the SAS Conn Module has been replaced).  I am going to look for a firmware update for the drive itself, which would be the only possible “fixable” issue.

More later

Posted in Uncategorized | Comments Off on It’s….it’s….ALIVE!

In search of the stable JS12

Well, the parts box didn’t remain unopened very long. We’ve tried a new SAS Connectivity Module and a new DSM.  The last round we tried the new HBA and at least have the JS12 back up.  The new fixpack was available this morning and it has been downloaded (4 CD’s worth) but I won’t be able to install them until I get back to the office tonight. The last few days *were* encouraging but after the attempt at a tape backup it seems that it didn’t take long before the wheels came off the bus and the blade fell silent again.

Was it the HBA?  I won’t know for a while.  Once the VIOS update is applied, I’ll try *several* tape backups to try to get the blade to fail.  We’ll see how it goes.  If the blog goes offline, you will know.

Posted in Uncategorized | Comments Off on In search of the stable JS12

If you can read this, the JS12 blade is running

Yeah.  It has been up and down. Still no answer as to why.  The blade has now been up for more than 24 hours and the only thing I have done since the nightmare began was to resume mirroring on the SAS disks assigned in IBM i which had inexplicably stopped AND I managed to get the LIC (RSE) restored and the latest CUME (TL09111).  So far it has been running like a champ.  It doesn’t add up.

Meanwhile, the box of parts that IBM sent today is still unopened.  Go figure.

Posted in Uncategorized | Comments Off on If you can read this, the JS12 blade is running

The i that isn’t

Where has the blog been the past 4 weeks? On inaccessible disks on a dead JS12 blade.  In fact, you may not be able to read this because the JS12 may be down again.  There is really no way to recount the difficulties of the past 4 weeks and the fact that my i partition has basically been DOA for a month.  There is a very long story here.  Too long to blog about in one sitting.

It all started innocently enough.  Just a couple of weird, unexplained situations where the JS12 would no longer be responding.  Bouncing the blade seemed to bring it back.  I decided to update the AMM with new firmware to see if it would “help” and then the slide down the slippery slope began.  Having a BCS vs an “integrated” i was part of the problem.  I finally activated the “phone home” option on the BCS, and when the i partition failed, the BCS phoned it in and I had a BCS support person call me.

The *NIX folks are MUCH more laid back than the i support group.  On the i side, on the very rare occasion that I had a failure, the support person would move heaven and earth to keep things moving until the i was back on it’s legs.  With the *NIX folks, it is pretty much “Try this, see if it works and call us back if it doesn’t”. Here is what that resulted in:

An eventual reload of VIOS

Update of the following firmware: Disk Storage Module (DSM),Storage controller, RAID controller, SAS connectivity module, disk firmware and a few others I have forgotten.  Each of these has dependencies and requirements to pay attention to.  Who would have guessed a controller update would lead to a disk firmware update.  When was the last time that you had to worry that a PTF on your storage controller would bugger your disks and require a separate maintenance piece?

The bottom line is that the i on blade is not an i (i being “integrated”).  It isn’t a criticism of  IBM per se, but you do need to consider the complexities in managing a blade environment.  There are definite benefits, but also real risks.  I’d take a look at the i on blade in 6 month to a year.  By then they will have sorted out these issues and perhaps come up with a more integrated way of manging the pieces.  Until then, you’ll have your work cut out for you.

As for my JS12, well, there is a new firmware update coming out next week.  Hopefully, I won’t have to update 6 other modules to make it work.  It is supposed to give me back some stability as well as allowing me to re-install the media tray so I can use the DVD drive again.  I’ll also be able to directly use my tape drive.  Imagine that! It’s, it’s almost like an i!

Posted in Uncategorized | Comments Off on The i that isn’t

Yes, WordPress runs on the IBM i

After a little experimenting I was able to get WordPress up and running on the i. The trick is to use the ProxyPreserveHost On directive in your httpd.conf file to get the proper url reported to WordPress.

I will continue to use Serendipity for my blog (also running on i, BTW) because I am lothe to convert the existing entries to WP format. But, were I to do it all again, I would use WordPress for my blog.

Posted in Uncategorized | Tagged | Comments Off on Yes, WordPress runs on the IBM i