gns3-server
gns3-server copied to clipboard
GNS3
Enhancement: Moving devices between remote servers
From gns3 community: https://gns3.com/discussions/moving-devices-between-remote-se
Moving devices between remote servers Discussion · General Networking I currently have a workstation as my GNS client and a GNS 3 VM running in vmware workstation as the main server. I also have 4 GNS3 VMs running on 4 different ESXi servers. Is there a way to move devices between servers?
I may be doing it wrong, but currently if I want to run the same device... like a VIRL switch on different GNS3 remote servers I have to create a template and manually specify a server. Now for a single device like the VIRL switch I have 5 different device templates each with a different naming convention.
We don't support it yet but it would indeed be nice. The difference device template workaround is the way to do it for now.
I'm sure you've heard it a 100 times. GNS3 rocks. Verbally describing a multihost environment on <insert your VM solution here> is much harder than pulling up the GNS3 gui and showing it.
I use GNS3 at my home to setup complex multi host environments. Nested ESXi servers hosting VMs along side other GNS3 devices such as iSCSI servers, ...
It can get complex which makes GNS3 sooo powerful.
The problem I have is that templates must specify the remote server to be run on. This means I cannot export my environment, ship it to another GNS3 users elsewhere, and have them import it in their environment.
I have 4 PCs that I use in GNS3. One master and 3 remote servers. Once I layout my machines, I'm stuck.
It would be great if as a template is deployed, I can select "use any available server" or "use any available remote server" or "use this specific server only".
This does require a rework of basic design of remote servers. That being said, it doesn't really feel like a huge request. Sure deployed template files will have to be migrated around. If a project is opend and closed, then they files have to rsync synchronized next time its started. Still this is not that hard. Another item that should be added is for devices to have a certain requirement. Say "only deploy me to a remote server that has 32 cores or 16G ram or both" It would of course need to support "your environment requires 67 cores and 384G ram having 13 sVMs, you cannot run this completely in this GNS3 environment as there are only 2 GNS3 servers with 32 cores and 128G ram, add more remote servers to this GNS3 environment".
It would also be great if the template were hypervisor agnostic. I know that's a big request. Delay that. I can live with this template only works in ESXi, or QEMU.
What would that give you?
- at home, bring up a 5th machine, next project restart, hosts are distributed across 5 machines instead of 4
- export a project that I had at home on 4 machines, copy to a remote GNS3 setup that only has 3 machines, bring it up just fine.
- grow your GNS3 environment by dynamically adding/removing remote servers.
- when testing, I build an environment, it's got everything in it (AD, DNS, DHCP, gateway, app server, sql server, test clients), I want to export this, ship it to the developer, and have them start it up even though their GNS3 environment has more remote servers than mine.
My imagination is failing me but this single feature would make GNS3 significantly more powerful than it is right now. This feature will add more use cases to GNS3.
Thanks,
I'm sure you've heard it a 100 times. GNS3 rocks. Verbally describing a multihost environment on <insert your VM solution here> is much harder than pulling up the GNS3 gui and showing it.
I use GNS3 at my home to setup complex multi host environments. Nested ESXi servers hosting VMs along side other GNS3 devices such as iSCSI servers, ...
It can get complex which makes GNS3 sooo powerful.
The problem I have is that templates must specify the remote server to be run on. This means I cannot export my environment, ship it to another GNS3 users elsewhere, and have them import it in their environment.
I have 4 PCs that I use in GNS3. One master and 3 remote servers. Once I layout my machines, I'm stuck.
It would be great if as a template is deployed, I can select "use any available server" or "use any available remote server" or "use this specific server only".
This does require a rework of basic design of remote servers. That being said, it doesn't really feel like a huge request. Sure deployed template files will have to be migrated around. If a project is opend and closed, then they files have to rsync synchronized next time its started. Still this is not that hard. Another item that should be added is for devices to have a certain requirement. Say "only deploy me to a remote server that has 32 cores or 16G ram or both" It would of course need to support "your environment requires 67 cores and 384G ram having 13 sVMs, you cannot run this completely in this GNS3 environment as there are only 2 GNS3 servers with 32 cores and 128G ram, add more remote servers to this GNS3 environment".
It would also be great if the template were hypervisor agnostic. I know that's a big request. Delay that. I can live with this template only works in ESXi, or QEMU.
What would that give you?
- at home, bring up a 5th machine, next project restart, hosts are distributed across 5 machines instead of 4
- export a project that I had at home on 4 machines, copy to a remote GNS3 setup that only has 3 machines, bring it up just fine.
- grow your GNS3 environment by dynamically adding/removing remote servers.
- when testing, I build an environment, it's got everything in it (AD, DNS, DHCP, gateway, app server, sql server, test clients), I want to export this, ship it to the developer, and have them start it up even though their GNS3 environment has more remote servers than mine.
My imagination is failing me but this single feature would make GNS3 significantly more powerful than it is right now. This feature will add more use cases to GNS3.
Thanks,
I think this is a duplicate of https://github.com/GNS3/gns3-server/issues/575
