Proxmox w/ gateway NFS and HA controller

LINBIT SDS Integrations Proxmox VE
, ,
1

Hi

I followed your tutorials to setup linstor on proxmox including controller HA using drbd-reactor.

However, looking now into addinf gateway for NFS:

https://linbit.com/blog/highly-available-nfs-for-proxmox-with-linstor-gateway/

you specifically mention that this tutorial is for starting with linstor without HA controller.

Is the combination making trouble or there any specific reason to not do both things at the same time (HA controller AND gateway for NFS ?).

Would be interesting to know.

Kr
M.

2

after reading further, still interested in feedback, though, is this about installing the gateway manager on all nodes, disabeling the service systemctrl disable linstor-gateway.service and adding to /etc/drbd-reactor.d/linstor_db.toml

[[promoter]]
id = "linstor_db"
[promoter.resources.linstor_db]
start = [
...
  "linstor-gateway.service"
]

to have drbd-reactor getting it auto-started on a node in the cluster?
As I tried that, seems to do the trick.

However, now one new issue showed up, can you look into this:

This is a 4-node cluster (node-1 .. node-4) no matter on which node the gateway is currently reunning, if I deplay nfs, it only is deployed on node-1, node-2 and node-4, but never on node-3. Is there any specific reason why node-3 is excluded?

I think this is the most relevant log I could find, but I can be wrong:

Mar 31 00:51:28 node-3 Controller[53947]: 2025-03-31 00:51:28.027 [grizzly-http-server-82] INFO  LINSTOR/Controller/e4df5a SYSTEM - REST/API RestClient(127.0.0.1; 'linstor-gateway/1.8.0-gdea45e8691ecb7da1d3c8c43fb1edd2d7ec75cc0')/LstExtFiles
Mar 31 00:51:28 node-3 Controller[53947]: 2025-03-31 00:51:28.030 [grizzly-http-server-83] INFO  LINSTOR/Controller/67d969 SYSTEM - REST/API RestClient(127.0.0.1; 'linstor-gateway/1.8.0-gdea45e8691ecb7da1d3c8c43fb1edd2d7ec75cc0')/DeployExtFile
Mar 31 00:51:28 node-3 Controller[53947]: 2025-03-31 00:51:28.066 [MainWorkerPool-20] INFO  LINSTOR/Controller/000047 SYSTEM - ACR: (node-1) Resource 'iso_storage' [DRBD] adjusted.
Mar 31 00:51:28 node-3 Controller[53947]: 2025-03-31 00:51:28.066 [MainWorkerPool-20] INFO  LINSTOR/Controller/000047 SYSTEM - ACR: Deployed /etc/drbd-reactor.d/linstor-gateway-nfs-iso_storage.toml on resource 'iso_storage' on 'node-1'
Mar 31 00:51:28 node-3 Controller[53947]: 2025-03-31 00:51:28.067 [MainWorkerPool-24] INFO  LINSTOR/Controller/000049 SYSTEM - ACR: (node-4) Resource 'iso_storage' [DRBD] adjusted.
Mar 31 00:51:28 node-3 Controller[53947]: 2025-03-31 00:51:28.067 [MainWorkerPool-26] INFO  LINSTOR/Controller/000049 SYSTEM - ACR: Deployed /etc/drbd-reactor.d/linstor-gateway-nfs-iso_storage.toml on resource 'iso_storage' on 'node-4'
Mar 31 00:51:28 node-3 Controller[53947]: 2025-03-31 00:51:28.076 [MainWorkerPool-31] INFO  LINSTOR/Controller/000049 SYSTEM - ACR: (node-2) Resource 'iso_storage' [DRBD] adjusted.
Mar 31 00:51:28 node-3 Controller[53947]: 2025-03-31 00:51:28.076 [MainWorkerPool-31] INFO  LINSTOR/Controller/000049 SYSTEM - ACR: Deployed /etc/drbd-reactor.d/linstor-gateway-nfs-iso_storage.toml on resource 'iso_storage' on 'node-2'

Thanks

Kr
M.

3

For the first question, it is permitted to use a HA controller setup with LINSTOR Gateway. You can see this outlined in the LINSTOR Gateway guide here:

As for why you are not seeing nfs shares being deployed on node-3, it is difficult to tell why that would be from the information shared, but you may wish to check that you have available storage there by running a linstor storage-pool list on the current LINSTOR controller node.