Doing an IOS upgrade

hurricane1091

So, we are going to be doing an IOS upgrade soon. I say we, but in reality the MSP is. So, I continue to get no experience and can't expand upon my base CCNA knowledge.

My question is though, what really needs to take place for an IOS upgrade? My understanding is that you would want to backup the image (.bin file) to a TFTP server, and then you can reverse the process to get the new IOS image onto the Cisco device. My question is in regards to the configs though. Is it necessary to back-up the running/startup config? Is that included in the process of backing-up of the IOS image? In the real world, is the running config usually the same as the startup config? In my labs I would save the running config to startup config at the end, just curious how the real world works. Also just curious what happens to the running-config after the IOS upgrade. My understanding is that the RC will go away since a reboot occurs and it's stored in RAM, but the start-up will remain (NVRAM) and so back to the original point, if the RC is the same as the SC everything should be okay.


Sorry if I should know this. I don't get to do CCNA stuff at my job even though it was listed on the job listing and something desirable. So it's hard to keep skills/knowledge fresh and understand how the real world works without ever seeing it.

RouteMyPacket

If you don't know the answer to this then why promote yourself as a CCNA R&S? So what good did getting the piece of paper do? Clearly nothing.

Sorry, just being realistic but alas I will also help answer your question.

There is nothing special about performing IOS upgrades. Depending on the platform being upgraded, you would want to read through the release notes to see what has changed, verify any known caveats etc. If there is nothing that would pose an issue to your specific environment then ensure a backup of all configs has taken place. If you have enough room in flash, upload the new image to flash along with the old image. Change the boot parameter to use the new image, reload the device and if all went well, you're done.

Rinse and repeat...

If you want to continue down the path of being a network engineer, ask yourself if your current position is assisting you in obtaining your goal/s. If the answer is no and it certainly sounds like it would be from what you have written, then what do you need to do about that? If a position is not fulfilling your desires and pushing your toward your goal/s then it's a waste of time and it's time to move on.

Invest in yourself, everything else is secondary. You getting your CCNA before you have had any experience is fine, to me it would show that you are interested in network technologies and in a face to face interview I would quickly be able to determine what kind of person you are and go from there.

You can practice upgrading switches at home and for now maybe ask yourself questions "How does inter-vlan routing really work?" and if you cannot answer it, then configure it and get it down...ask yourself another question and onward and onward. This way you are learning and becoming more confident in being able to apply those skills, being able to do it in the comfort of your home is nice.

Again, you have to invest in yourself and constantly push forward, it's too easy to become complacent.

hurricane1091

Can you revisit what I said and point out what isn't true before trying to say the CCNA did not help? Because if everything I said is true (startup-config being stored in the NVRAM for example) then why would you say that? Real world example and what I do in my room are completely different. You don't know that I haven't done an IOS upgrade at home and not worried about backups because there was no need to, so why would you make assumptions like that?

You don't know my life, or my journey. I tried to break into the networking field, and it did not happen. So I am getting a BS degree to add to the resume to help get into the field. I WILL study & practice the CCNA R&S material again in a year when I get ready to try to break back into the field again. In the mean time, I try to keep fresh the best I can but it's already been awhile since I passed the exam, I will naturally forget things if I do not get to put them to use at work (and am studying for a degree at home).

RouteMyPacket

My question is though, what really needs to take place for an IOS upgrade?


My understanding is that you would want to backup the image (.bin file) to a TFTP server, and then you can reverse the process to get the new IOS image onto the Cisco device.


Sure, sounds good


My question is in regards to the configs though. Is it necessary to back-up the running/startup config?


Nah, running-configs aren't important. /sarcasm Of course back the config up first!!!


Is that included in the process of backing-up of the IOS image?


No, you have to do this yourself.



In the real world, is the running config usually the same as the startup config?


I don't know, is it? Have any changes been made and not saved? If so, then yes startup and running will be different.


In my labs I would save the running config to startup config at the end, just curious how the real world works.


Real world, bedroom, strip club, bathroom, it doesn't matter. Save the config before upgrade


Also just curious what happens to the running-config after the IOS upgrade.


Depends on the platform and current configuration in place, this is why you would read the release notes for changes and caveats. Some configurations might be deprecated etc.

hurricane1091

RouteMyPacket wrote: »

Is that included in the process of backing-up of the IOS image?


No, you have to do this yourself.

This is the information I needed answered, thanks. I was thinking that you needed to do it yourself because if you loaded the new IOS up and the startup config was not there for some reason, having the startup config in the old IOS image would not be helpful anyways.

Always back things up, yes. It sounds silly asking if it was necessary to really back up the configs, but that was because I did not know if it was included in the IOS image backup. Since I know it's not now, then obviously it's absolutely important to back it up! Even if it was, it wouldn't be helpful because of the reasoning I used above!

And I know it sounds stupid asking if the running-config is always the same as the start-up config. I could go into packet tracer and set up a simple OSPF network, and never save it to start-up config and everything will be fine unless I shut down the router. I would assume that the running-config matches the start-up config always, but how would I know if there was some sort of situation in the real-world where that isn't the case?

I really hope you can see where I'm coming from and that I actually have an understanding of things. I remember learning about server 2012 in college and then got to the real world and found 2003. I was like "wow that's so old!" but quickly learned and understood why businesses are hesitant to upgrade and the work that needs to go into that to make it happen. You don't learn that in college you know? Just like some things you do not learn from a certification.

RouteMyPacket

hurricane1091 wrote: »

I would assume that the running-config matches the start-up config always, but how would I know if there was some sort of situation in the real-world where that isn't the case?

NEVER assume anything. Unless you are literally the only Engineer in the environment with access to your gear. Otherwise, never trust anyone else's word, you bounce that device and for whatever reason it doesn't come back up with the old running config, who do you think get's the blame? Yeah, YOU! So, again...NEVER assume...too many tards out there.

When I go into an Enterprise and need to do something, regardless of what it is. I have made it a best practice to immediately take a backup of the config, advise the client. Save the config as is, make my changes and go from there

hurricane1091

RouteMyPacket wrote: »

NEVER assume anything. Unless you are literally the only Engineer in the environment with access to your gear. Otherwise, never trust anyone else's word, you bounce that device and for whatever reason it doesn't come back up with the old running config, who do you think get's the blame? Yeah, YOU! So, again...NEVER assume...too many tards out there.

When I go into an Enterprise and need to do something, regardless of what it is. I have made it a best practice to immediately take a backup of the config, advise the client. Save the config as is, make my changes and go from there

That sounds good. And that makes sense. I'm glad to hear what you do in the real-world. I really would like to get there one day. I know I would need to brush up on everything before trying again to enter the field, and I will. It's tough though, the role I'm in is a Jr. Systems Analyst and the MCSA would be very helpful over the CCNA. I can't help it though, I thought my interest was also in the MCSA/MCSE realm but it's starting to become clear that it isn't. And I failed to get a Jr. Network Engineer role while the material was fresh out of the oven, and decided that the best course of action was to continue to get IT experience since I'm very young and finish my BS since it shouldn't take over a year, and then try again to enter it. I couldn't set you up an EIGRP network today, I'll admit that. The commands will get me. But I know what it is, I know how it works, and with some command googling/notebook help I could get it going, I promise that. That wouldn't be good enough if I was actively interviewing for a position in the field, but for right now I think it is. Do I know what companies prefer in the real world though? No. I just know we do not use it here.

RouteMyPacket

Ok, then as I said before; ask yourself that question "How do I configure EIGRP between two devices?"...then go and do it

Cisco has everything you need, being able to google search is a skill in and of itself

Cisco IOS Software Releases 12.4 Mainline - Configuration Guides - Cisco

Then look into IP Routing and there it is

IP Routing: EIGRP Configuration Guide, Cisco IOS Release 12.4 - Cisco

clarson

well here is how the real world works. losing data is like dropping the ball in football. if you drop the ball, nobody is going to trust giving you the ball. So, you better play a position that doesn't touch the ball or nobody is going to want you on their team. So, all you have to ask yourself is will this lose data. having a running config that is different than your starting config will lose data. you lose data if the router fries and it isn't backed up. you lose data if you all your backups are on site and you lose the site. how much risk is acceptable is dependant on how much downtime is acceptable. but, if you lose cheap data, no one is going to trust you with the expensive data. You show up late for meetings, nobody is going to trust you finish a project on time. just the way people are.

Heero

1. write the config
2. Backup the config to another device.
3. If you have room in flash, tftp the new IOS to the router's flash. If you don't, you have to delete the old IOS first and then TFTP the new one. If the router dies after you deleted the currently running IOS from flash but before the new one is completely copied over, you get to do a fun xmodem recovery.
4. make sure the boot variables are set correct for your new IOS
5. reload and verify everything looks good.

Nightflier101BL

I performed an IOS upgrade on one our core switches not long ago. While everything worked out fine, I was thinking what I would do if the image was corrupt and the switch didn't boot the new IOS. In this case, we could fit the new image along with the old in flash, so not that big of deal. But what if it was only the one image?

Just to see how it works, I took an old 2950 and deleted the IOS and rebooted. I used HyperTerminal and xmodem to get the old deleted image back onto the switch (after spending hours trying to figure out how to use iTerm on a Mac to do this - ended up using a Windows laptop). I would not have wanted to figure all that out with a paperweight core switch and an entire building screaming at me.

Always backup everything. Whenever you go to do something, ask yourself "what am I going to do if this doesn't work?"

JeanM

Also in real world, you wouldn't do a maint. update in the middle of business hours right ?

There should be a) redundancy in place b) change management in place c) maint. schedule.

Nightflier101BL

Not if I can help it! Unfortunately, this was for a 24/7 regional jail facility.