Stability of beegfs running in mixed Rocky 9.5 and 9.6 environment

[Faraz Hussain]

I have taken over a small 10-node cluster which is using 3 file servers running beegfs 8.0.1 .  Some nodes are running Rocky Linux 9.5, others 9.6. The ones running 9.6 the beegfs would not work, so I had to install the latest MLNX OFED drivers for 9.6. Then they worked.

Is beegfs stable running as it is now or am I risking corrupting the file system? I'd like to upgrade all machines to Rocky 9.6 but I first want to better understand how beegfs works. Here's the output of beegfs health net below. Is it normal that the management node connection is ethernet instead of rdma?

---------------

Management Node

---------------

management [ID: 1]

   Connections: ethernet: 1 (172.30.190.17:8008);

--------------

Metadata Nodes

--------------

node_meta_1 [ID: 1]

   Connections: rdma: 2 (172.30.194.17:8005);

node_meta_2 [ID: 2]

   Connections: rdma: 1 (172.30.194.18:8005);

node_meta_3 [ID: 3]

   Connections: rdma: 1 (172.30.194.19:8005);

-------------

Storage Nodes

-------------

node_storage_1 [ID: 1]

   Connections: rdma: 1 (172.30.194.17:8003);

node_storage_2 [ID: 2]

   Connections: rdma: 1 (172.30.194.18:8003);

node_storage_3 [ID: 3]

   Connections: rdma: 1 (172.30.194.19:8003);

[Denis Anjos]

Hello, Faraz,

 

As for the management node connection, it is normal its normal behaviour. It  only speaks TCP.

D.

 

--
You received this message because you are subscribed to the Google Groups "beegfs-user" group.
To unsubscribe from this group and stop receiving emails from it, send an email to fhgfs-user+...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/fhgfs-user/1eaf61fe-bb76-4e69-ab07-70a885b473a0n%40googlegroups.com.