John H Goodwin

[@JonathanD] Todays objective is to demonstrate the management module in the IBM Bladecenter.

[@JonathanD] you'll note I expanded yesterdays array from 3 drives to 5

[@JonathanD] To show the functions and features that can help you manage a chassis and the blades.

[@JonathanD] everyone connected?

[@JonathanD] we're going to step through top to bottom

[Attendee1] Currently, this is the RSA for the blade?

[@JonathanD] This is the Management module interface. it is very similar to what an RSA provides

[@JonathanD] but there is only one for all blades.

[Attendee1] Oh, right, just looks like the RSA interface.

[@JonathanD] as you can see, there are 2 blades installed

[@JonathanD] one of them has a sidecar, making it take 2 slots.

[@JonathanD] right now we're looking at the "system status" page

[@JonathanD] which gives a high level view of the health of the chassis, blades, and modules.

[@JonathanD] Next, we have the event log

[@JonathanD] To give us something to look at, I'm going to cause a failure.

[@JonathanD] One moment

[@JonathanD] and theres a message

[@JonathanD] Insufficient chassis power to support redundancy

[@JonathanD] and you can see the error where I actually removed the power module

[@JonathanD] also, which module was removed

[@JonathanD] in the event of a failure, you would be able to identify which one was failed, from here

[@JonathanD] back on the system status page, you can see we now have an indicator at the top telling us of a failure

[@JonathanD] and here, the missing power supply

[@JonathanD] any questions so far?

[@JonathanD] and the module is back

[@JonathanD] next, we have system LEDs

[@JonathanD] here you can see the state and turn certain LEDs on and off

[@JonathanD] specifically, the location light, which is blue. This feature is useful in hosted datacenter or if you must direct someone to specific server

[@JonathanD] right now, I have the location led on my LS21 blade flashing

[@JonathanD] and I can indeed see it from here :)

[@JonathanD] next, we have power availability and consumption

[@JonathanD] the "fuel gauge"

[Attendee1] Consumption is nice to calculate costs of runtime.

[@JonathanD] this shows how much power we are using, how much is available, etc.

[Attendee1] Power in use, is that with a reserve amount, or actual draw?

[@JonathanD] Attendee1: these numbers are actually not 100% accurate, they are ALL higher than the "real" draw. There is a way to get down to the watt, per server and device, and I think I will demo that sometime next week.

[@JonathanD] This is done via IBM PowerExecutive, which is free for IBM servers.

[@JonathanD] allocated power is generally speaking the actual draw

[@JonathanD] although really, it's the possible consumption in peak conditions.

[@JonathanD] Next, we have the hardware VPD, or vital product data

[@JonathanD] here, we have a listing of the model and serial of all the major components of this system.

[@JonathanD] this can be very useful if you ever have to call for support.

[@JonathanD] here, we have the firmware VPD, giving us firmware levels on the various components in the Blade chassi

[@JonathanD] as you can see, even the fans have a listed firmware level :)

[@JonathanD] we're going on to blades now :)

[Attendee1] Okay, when you install blades, should you space them while you have room, or does it matter?

[@JonathanD] it doesn't matter, with one possible exception

[@JonathanD] if you purchase a chassis with 4 PSU (you don't have to day 1, you can get 2, and add 2 later)

[@JonathanD] you might want to put half the blades in power domain 1 (slot 1 to 7 or 1 to 6, depending on which chassis)

[@JonathanD] and half in domain 2, which includes the remaining slots

[@JonathanD] to maximize availability

[@JonathanD] we will generally build for failures that may well never happen, but IF you have the capability to, essentially for free, you might as well take advantage of it.

[@JonathanD] Like the enclosure loss protection we talked about yesterday.

[Attendee1] Right. If you have it anyway, should go ahead and use it, unless you expect to downsize and don't want to rely on it.

[@JonathanD] correct

[@JonathanD] so, on this page, we can turn blades on, off and restart them.

[Attendee1] Looks like you can do a batch of them?

[@JonathanD] similar to the RSA, but for any blade in the chassis, from one point

[@JonathanD] yes

[Attendee1] Does it offset the power-ons?

[@JonathanD] yes, but only if you select more than a couple

[Attendee1] Oh, just 2 it won't then.

[@JonathanD] Yes, if you select 2, they'll both power on more or less at once.

[Attendee1] But, all 14, it would.

[@JonathanD] Yes.

[@JonathanD] we can also restart the system management processor

[@JonathanD] which is what provides this interface into a blade to begin with.

[@JonathanD] we're going to skip on demand.

[@JonathanD] but basically, you can keep a blade onsite, and only pay for the time you use it.

[@JonathanD] here, we have remote control

[@JonathanD] I'm going to start a remote control session now.

[@JonathanD] now, we have this fun little "remote disk" mount utility

[@JonathanD] which is very useful.

[Attendee1] I noticed the chassis has a single optical, and floppy drive.

[@JonathanD] it basically allows you to mount your local cdrom or an iso, on the blade, from your desk

[Attendee1] How does this get shared among the blades?

[@JonathanD] yes, but the H chassis has no floppy

[Attendee1] HS20 has a floppy doesn't it?

[@JonathanD] Attendee1: it's owned by 1 at a time. There is a button on the front of each blade to switch

[@JonathanD] Attendee1: the enterprise chassis has a dvd and floppy

[@JonathanD] the H has only a DVD

[@JonathanD] none of the blades, themselves, have either.

[@JonathanD] Attendee1: HS20 is a blade, specifically

[@JonathanD] you can also share the media tray amongst the blades, instead of switching it

[Attendee1] I like the idea of shrinking the headings on the RSA.

[@JonathanD] this makes for slow access times, though.

[Attendee1] My older card doesn't have that option.

[@JonathanD] Attendee1: yes, it's more effective of course in a larger display.

[@JonathanD] here, we have blade configuration

[@JonathanD] one of the most interesting features here is the boot order editor, which lets yo umodify boot order without entering the bios on each blade

[@JonathanD] rebooting, hopefully to our mounted iso

[Attendee1] Xeon 51xx series. I'm so envious.

[@JonathanD] they're peppy little things, I gotta say.

[@JonathanD] and there it is

[@JonathanD] the IO modules are a bit outside the scope of todays discussions.

[@JonathanD] so we will skip that, for now.

[Attendee1] MM stands for what?

[@JonathanD] management module

[Attendee1] Okay, so this is to control the management module itself.

[@JonathanD] right

[@JonathanD] here, we have login profiles, I have only one (the default) defined

[@JonathanD] you can assign different rights to each login. lets take a look at that.

[@JonathanD] we'll give this account rights to the chassis, and to the HS21 blade

[@JonathanD] and there we are

[@JonathanD] bob is created

[@JonathanD] next, we'll look at alerts

[@JonathanD] I have one alert setup, going to IBM Director, a free, IBM server management package

[@JonathanD] we can create additional alert recipients here

[@JonathanD] and control what they will be sent here.

[@JonathanD] the blade chassis supports ssl for a secure connection, which can be configured here.

[@JonathanD] here, we can update the firmware of the MM.

[@JonathanD] this is non-disruptive, all blades stay online.

[@JonathanD] finally, once your chassis is setup properly, you can backup the configuration to the chassis itself, or to a local file.

[@JonathanD] this is useful if you ever must replace the management module, where all this information is stored.

[@JonathanD] and that about sums it up

[@JonathanD] since we're running out of time here, if there are any questions?

[@JonathanD] I guess thats about it for today then.

[@JonathanD] Tune in next week for more exciting adventures of "SuperBlade"