Would like to get a view from you all...
I was thinking of upgrading the IOS of a number of c3750 stack (roughtly about 50-100 stacks around the country)...I would like to know is this feasible?
My CiscoWork NMS is connected with a low bandwidth (2Mbps) management link. I was thinking if I upgrade via CW2k, then RME will take "ages" to push the IOS to the stack, right? Is this feasbile for 100+ stack of switches?
Yes upgrade of stacks is possible. If you have low bandwidth links, try to use the Remote Staging feature in RME.
The Remote Staging and Distribution option helps you to upgrade multiple devices over a WAN.
You can perform remote staging and distribution by either of the methods:
•External FTP server
•External TFTP server
•Remote Staging Device
In this workflow, a managed device or external TFTP server or FTP server is used to stage an image temporarily. The staged image is then used to upgrade devices that are connected by LAN to the Remote Stage device or external TFTP server or FTP server. If you use this method, you do not have to copy a similar image, multiple times across the WAN.
After the image distribution job is completed using a managed device as a remote stage device, the configuration changes made to the Remote Stage device are automatically reversed and the staged image is deleted from the Remote Stage device.
For more details check :
Thanks.....Like to know what happen if the flash in the switches has no enough space?...Is the RME smart enough to backup the old image; delete the old image; upgrade with the new image?
Yes, it is. Reference.
"During the image upgrade, Software Management:
•Checks the amount of Flash memory on the device. If Flash memory needs to be erased before the new system image is loaded and erasing is allowed, it erases the Flash memory. Before erasing the Flash, a warning message appears
Flash memory will be erased."
I'd also run an upgrade analysis job on your target stacks first. Even when it worksd properly, a 3750 stack can take a long time to upgrade since the image is copied from the stack master and applied to every member.
But what happen if the upgrade fail and the stack was not able to boot up? Can RME put back the old IOS and boot it up? Or we need some engr. to go onsite to console in and do the upgrade from there?
If yes, then it defeat the purpose of having RME to perform a remote upgrade? Thoughts?
Ummm. If a device (or stack) won't boot following a failed upgrade attempt no NMS in the world will fix that. Without some sort of running conifguration loaded you won't be able to reach the device other than on console.
LMS will allow you to manage your images, give you awareness of the deployment state across your enterprise, analyze your readiness to deploy upgrades, manage the deployment sets (optionally incorporating role-based workflow management), notify you of the status of your upgrade jobs, etc. etc.