350-601 DCCOR Cisco CCIE Data Center – Compute part 9

28. UCS Firmware upgrade 02

Let us continue and check other restaurants as well. So we have all point number one to nine and one by one we are going to exploring that. Now. In this example we will go and check that from Intersect, how we can go and get the references? How we can go and get the notifications that you have these issues you can acknowledge and then you can proceed for your upgrade, correct? So what you need to do simply go to interstate and we have complete session over Intersight in prior sections. It’s very easy to you can go to the dashboard servers once you’re inside the server then you can go and check the health. So here you can see that it is clearly showing that health is critical or healthy. And then you can go and see that the HCL status correct. Now once we found this, then we can go and check the recommendations as well.

So inside the server you can see that we have this option that get the recommendation and we can go and download the drivers and firmware. Now, it’s very important that prior to do the upgrade you should have the up to dated firmware. Correct? And that’s one of the recommendation I will show in the upcoming slide. It’s very easy that you can go and you can download the recommended drivers and then once you have these recommended drivers, you can go and do the installation. Correct? But step number you can say that four is that you should have the recommended tools, the hardware so before doing the upgrade you can see now we have completed step number one that the notification then you can go and use the HCL tool then we should go and check the pre upgrade checks.

Very important again what are the pre upgrade checks we have and here you can see that we have a complete list so you should go and check the outdoor configuration backup if you have any critical major faults fi reboot if it is pending you can go and check the entity configuration as well. Now, very important point that your available boot flash space should not be actually should not be more than 70% means you should have empty space. 30%. Correct. Then we can go and check different type of failover. So either it’s a VHV failover. Either. It’s a vnik failover. We should go and check these things. Okay, again you can see that when we have the partition, you can see that you should not overcongest the partitions. Correct. You should have at least 30% space there in case while doing the upgrade, if the operating system will grow, then your system will not go and crash.

So you can see here that step number four is that prepare for firmware install. That means once you have all the pre checks, then you should go and check that what type of firmwares you want prior to actual upgrade, right? Let’s go and explore this option as well. Prepare for firmware install and again you can see that we have option. I told you earlier, we have option.You want to do the upgrade via UCSM, via UCS Central, via Intersight. You want to do manually, you want to do auto install. So auto install options are there, manuals are also there. If someone wants to do manual, so that is also there. But before doing the upgrade, again we have this firmware install. What you can do, you can go to UCS manager equipment, you can go to the firmware management and then you have this firmware or to install.

Prepare firmware for firmware install means you are ready for the firmware install. Why this step is important because before doing the actual upgrade, you should have the readiness, that’s the key, readiness. So your actual upgrade is ready. Have you checklist all the readiness parameters and that’s the main reason we are doing this firmware install, correct? So you can see configure the notification, use online HCL tools. There are so many nice tools. You can go and check, do the pre upgrade checks, then prepare the firmware install. And then the fifth step here we have in the list is use the fault separation. And this is also very important. We should go and check what type of faults are there before doing the upgrade. And we have the option that we can go and do the separation start fault separation.

You can see that we can go to the equipment and inside equipment, you can see that chassis server, IO models inside the equipment general category, you’ll find that we have options related to do the fault separation, related to do separation task properties, correct? So you can see clearly and it’s very much west that mute button for UCS error states system maintenance will not raise alarm to help disk can specify a time or completion of activity done on server chassis, IOM level. So you can see it’s done on chassis label. It’s done on server level, it’s done on IO label. Okay, so we can go and do the fault suppression during the migration. Now the suppression mechanism, and this is one of the pictorial representation here you can see that the UCS manager is getting the events and again different pieces inside the UCS manager, correct? So we have that option, correct?

We can go here and we can do the fault separation related to chassis servers and the I O model. So it’s okay. Again, if you want to check that, where is what you can go and refer this particular diagram, correct? So UCS manager is there and then he’s taking all the pieces. He’s doing the false separation. So you can see that false separation. And again you can see the different type of monitoring engines that we have. Either is IBM based or HP based or SNMP. Everything is like, okay, stop all these notifications because we are going to do the migration. We are going to be ready for the migration. All right, so we have completed up to a step number five, and we can start with a step number six and the rest of the steps in the upcoming section. All right? So let’s stop here.

29. UCS Firmware upgrade 03

Let us continue for two more videos. So here I’m going to hit point number six, seven and eight. First of all, we have to use the baseline policy. What is the agenda here? So we should know that before upgrade, what are the errors, what are the issues we have? And then after the upgrade, what are the issues and errors are coming. So that’s the thing that we should notify the alerts and alarms before the upgrade and after the upgrade. And for that we can go and create the baseline policy. So while creating the baseline policy, this is some sort of marker that pre and post upgrade. What are the alarms and alerts we have that we should go and address? Correct. And that’s why you can go and give the baseline expiration interval as one day just to make sure that you understood that the pre and post alarm and alerts. Correct. The next step is very important, very critical.

That is the backup. Now we have options that we can take full backup. And it’s recommended that you can go and take the all configuration backup. You should go and take actually full backup. Correct. And now once you are taking the backup, you have options. Before the upgrade you can take the backup R, you have auto backup enable. So you have your SFTP server where you are doing this auto backup, maybe daily or maybe weekly. It’s up to us how important that backup policy we have in our organization. Some of the organization, they are taking it daily, some they are taking weekly or biweekly, etcetera. So as for the backup policy, we can go and get these backups or do this backup with respect to secure protocol such as Sfkp. Correct. Again, one advantage we are doing auto backup is that if you have any catastrophic failure, that means you have the backup.

So you’re not worried because you have some backup you are taking. As for your backup policy, you are taking the backup. So if any failure is also there, that is unknown failure, et cetera. So we can recover from the backup. Now, here you can see two auto thing setting the auto backup. It’s very easy. We can go and set like this, you can see the policy backup and export. And then we have one option that automatic UCSM backup during auto install. So while doing auto install as well, we can set this auto. So three backup things are there. You can take the backup before your actual upgrade.You can create some backup policy for auto install or during the installation as well. We can go and take the backup as the auto backup.

So here you can see that perform the auto automatic fullest state backup during the firmware auto install operation and performs best practice, fullest state backup, fullest at backup of system and you have this backup file as TGZ. Great. Then again, this is again very important. Step number eight that you use the fabric evocuation methodology. Why? Because you don’t want to disturb the traffic. Correct. So what does it mean? Let me first of all explain to you about this Evocation, what exactly it is doing and then we’ll come and discuss little bit more about that. So you can see here that while in the normal operation you can see that our traffic is load balanced and it is using both the uplink FIA and fib and it is going from south to north.

Correct? Now when you’re doing the upgrade, half of your system is shut down for example or it is an upgrade or maintenance mode, then your traffic, so your Vneck or your VHP traffic should be pinned or should be properly load balanced towards the other uplink, towards the other fi, right? And we should do the configuration related to that and that’s with the help of proper evacuation method. So while we are doing the upgrade and we don’t want the disruption, we want minimum disruption so at that time we can go and configure the Evac mode. So Evac mode is on, that means it will go and check the traffic load balancing from one for example Venice to other VENIC or from one fabric A to fabric B. Like that we can think correct.So that’s the evacuation method we should use and while we are doing that we can see the normal operation and the featured operation of feature operation.

So in that case you can see the admin evacu mode is on, operational Evac mode, evacuation mode is on. Great. And how we can do that while doing the auto install so while doing the auto install also when you are doing that, when you go and open the install infrastructure firmware workflow, at that time you will get the Evacuete option. Let me show you that. So you can see that we are doing the install infrastructure firmware in that we have a step number two in the workflow install infrastructure firmware where we have the Evacueate option that we can go and check mark okay, great. So we have completed up to point number eight in next section what I want to do, I’ll go and do the summarization. So we have one point left, we’ll discuss about that you see us or to install a lightweight patching and then we’ll are good to complete this particular section.

30. UCS Firmware upgrade 04

While doing the upgrade, we have option. So we know that we have option and we have discussed about this earlier. We want to do manual, we want to do automatically, we want to do via the UCS manager center or Intersight correct. Now let’s quickly check the manual versus auto upgrade process. Now, if we are doing manual upgrade rate, that means that we have to check mark each and every step correct. So we have to start say for example, activate the UCSM, check if UCSM upgrade validation passed the UCSM upgrade as the expected version, check the IOM, check the FIAs, check the subordinate inside FIAs, check the user acknowledgement, activate the primary and then go for the failover, etc. That means we have to do all these steps, this checklist one by one, step by step. This is the thing that when we are doing the manual firmware upgrade then the other flavor that we have is auto install.

Now, while we are doing the auto install, that means that UCS is giving us one workflow and again, you know that when we are creating the service templates, when we are creating the service profiles, you know that we have the workflow in UCS you have the workflow lines workflows. We have that if you go and follow those workflow, then that will ensure that whatever you are doing step by step, those things will come inside the workflow. Not as exactly that you have in the manual steps, but they have their own defined steps. So while we are doing the auto install, you can see that it will go and check the previous site, then it will check the package version, then you can go and select the firmware package, then dependencies and then it will go and do the upgrade correct, means you want to do the upgrade.

 So you can go and you can follow the infrastructure firmware server firmware like that. One nice thing inside auto upgrade is this that the auto upgrade nowadays they are understanding the software versus the hardware. So that means they will go internally and do the hardware enablement for software plus they will go and do the firmware upgrade as well. That’s a good thing that we need not worry about how the things behind the scene they are upgrading. So we know that we have tight and high end hardware and those hardware where you have the NX OS operating systems and other firmware software and packages correct. So the auto upgrade will go and take care about that. All right, so this is related to these sections, related to a UCS firmware software upgrade and we have completed this section. So we are good. We can stop here now.