AWS CHR Upgrade to 7.3.1 Fails to boot LoaE01

RouterOS General
1

I have a CHR level P10 install on an AWS EC2 instance type c6i.large that is working currently on version 7.1.2. Check for updates, channel stable, current version 7.1.2, latest version 7.3.1, click download and install. Instance then fails to boot. I’ve attached a screen shot of the console. Booting from Hard Disk 0… LoaE01

I’m able to recover to instance by replacing the root volume with a snapshot I took before the upgrade. I’ve tried it a couple of times with the same result each time.

The instance has 1gb of gp3 storage with 968MiB free. The CHR instance was created long ago using the AMI created by MikroTik.

I hope someone has an idea on how to upgrade this CHR.
LoaE01.jpg

v7.13.5 [stable] is released!
2

I also just tried upgrading to the development channel 7.4rc2 and it has the same failure. LoaE01 on the console.

3

The same thing just happened to me a couple of minutes ago.

4

I figured out some steps that will hopefully help MikroTik resolve the issue:
Subscribe to MikroTik Cloud Hosted Router in the AWS Market Place and launch 1gb gp3 storage, c5.large instance type
Starts out at version 6.44.3
Download and Install Upgrade RouterOS to 6.49.6
Download and Install Upgrade RouterOS to 7.3.1
Uninstall dude package and reboot
Now it’s broken, won’t boot, LoaE01 on the console
i-0c7e3f355b52ae26d.jpg
Is the dude package required on CHR?

5

Dude is not required but I think it’s probably time to kill it off and spin up a new one. Have you reached out to MT support on this?

6

I have the same problem,
Any solution ?

7

Export (remember show-sensitive) the config, start a new one from fresh disk image, import your config.

8

I’ve been dealing with this same issue and working with Mikrotik support since March. Previously I was unable to figure out how to create my own AMI. This weekend I finally realized what I was previously doing wrong. I have documented in relative detail the issues I’ve experienced with T3/T3a instances, and how to create your own 7.6 AMI so you can create stable ROS 7 instances.

Creating a CHR Instance in AWS EC2

I’ve provided this information to Mikrotik support (ticket SUP-77812). Hopefully they’ll create an ROS 7 AMI soon. In the meantime, you have the information you need to create your own.

9

I just had a similar issue happen with two of my CHRs in AWS. Remove the dude package and reboot, then the instance won’t start.

Instance state “Running”. Status check “Initializing”, can’t open an E2 serial console. It appears the instance just won’t start.


Robert

10

Hi, thanks for the guide in you site… such easy steps and MikroTik “can’t” simply update the CHR version in their AWS Marketplace…

Question, in your testings, are you using GP2 or GP3 EBS Volumes?

11

You can un-install the dude before updating to final v6

Then Update to final v6, and update to latest v7

Dont try to add the dude package after, it bricks the CHR instance, wont boot at all. Not even a LoaE01 error.

12

Hi,

This is only a problem when using AWS T3 instances. T2 does not have this issue.
Also, you can change the instance to t2, update, change back to t3 and all working.

I have a ticket open on the issue, Support thinks this is an issue with the t3 hypervisor.

v7.13.5 [stable] is released!
13

V7 hasn’t been upgradable from the beginning on T3, great thanks for T2<->T3 trick, will try it soon.