Posted on March 22, 2017 7:34 pm
 |  Asked by Lucas Albuquerque
 |  1478 views
Tags:
0
0
Print Friendly, PDF & Email

Hi,

I’m trying to run Arista 4.18.0 (aboot 8.0.0) on GNS3, locally, but i’m having issues to initialize the device.

I followed the setup on the link https://eos.arista.com/running-veos-in-gns3-1-5/, changing only the part about the server, which is local. All good until i turn the device on. It takes a lot of time to boot before i decided to stop it, and here what i got from the console:

Aboot 8.0.0-3255441
running dosfsck
Seek to -512:Invalid argument
dosfsck 2.11, 12 Mar 2005, FAT32, LFN

Press Control-C now to enter Aboot shell
Booting flash:vEOS-lab.swi
[ 13.899845] Starting new kernel
[ 4.228263] Running dosfsck on: /mnt/flash
fsck.fat 3.0.23 (2013-10-15)
/dev/sdb: 4 files, 71388/1046527 clusters
Data in /mnt/flash/vEOS-lab.swi differs from previous boot image on /mnt/flash.
Saving new boot image to /mnt/flash…

Switching rootfs

Welcome to Arista Networks EOS 4.18.0F
New seat seat0.
[ 299.050688] systemd[1]: Failed to start Login Service.
New seat seat0.
[ OK ] TimeAgent: Starting ConnMgr: [ OK ]
Cannot find initial response [FAILED]
Starting ProcMgr: [ 389.480340] systemd[1]: Failed to start Login Service.
New seat seat0.
Failed to start Login Service.[ 480.114004] systemd[1]: Failed to start Login Service.
New seat seat0.
[FAILED] Starting EOS initialization stage 1: [ 534.793421] NMI watchdog: failed to be enabled on some cpus
[ OK ] Starting NorCal initialization: [ 570.571657] systemd[1]: Failed to start Login Service.
New seat seat0.
[ 661.233323] systemd[1]: Failed to start Login Service.
New seat seat0.
[ OK ]
Starting EOS initialization stage 2: [ 751.714978] systemd[1]: Failed to start Login Service.
New seat seat0.
[ OK ]

Starting Power On Self Test (POST): [ OK ]
Completing EOS initialization (press ESC to skip): [ 842.432648] systemd[1]: Failed to start Login Service.
New seat seat0.

[ 933.390332] systemd[1]: Failed to start Login Service.
New seat seat0.

What am i doing wrong?

 

Thanks in advance!

 

Regards,

 

Lucas.

1
Posted by Tamas Plugor
Answered on March 24, 2017 4:28 pm

Hey Lucas,

You’re not doing anything wrong technically. The problem is that you are using the local server, and qemu implementation for windows is extremely slow compared to linux, so it is recommended to use the GSN3 VM. I did a couple of tests, and the switch will boot eventually if you wait long enough (around 15-25 minutes), but it’s still running really slow.  VMware Player should be enough for the GNS3 VM if it’s not possible for you to get a copy of VMware workstation. Another recommendation would be to use EVE/unetlab, you can run it on a VM, and it’s compatible with a lot of vendor images, including Arista and it runs smoother than GNS3.

Let us know if you have more questions.

Thanks,

Tamas

Marked as spam

Hi Tamas,

I have installed it on VMWare ESXi 6.0. Problem is when I turn on vEOS i get this error

LZMA daa is corrupt
Loading initrd…….ready.
— System halted —
Aboot 8.0.0-3255441
running dosfsck
Seek to -331320832: Invalid argument

Booting flash://vEOS-lab.swi

Then i can see nothing

Do you happen to know how to tackle such issue?

Thank you
AP

(Arpit Patel at April 19, 2018 5:37 pm)
0
Posted by Tamas Plugor
Answered on April 19, 2018 9:06 pm

Hi Arpit,

It might be because of the provisioning issue when using ESXI 6, in which case, you’ll need to convert the vmdk to be thick provisioned.

The following article should be helpful: 

https://eos.arista.com/tip-for-arista-veos-on-vmware-esx-6/

The pre-requisites from the following article mentions the same:

https://eos.arista.com/running-veos-on-esxi-5-5/

Hope this helps!

Thanks,

Tamas

 

Post your Answer

You must be logged in to post an answer.