Hi all,
I'm testing beegfs inside virtual machines. My first attempt with ubuntu 16.04 was successful but with centos (RockyLinux 8.4) the metadata node gets no answer from management. This is straightforward configuratio following the quickstart guide.
ctrl0 $ cat /var/log/beegfs-mgmtd.log
(1) Sep07 20:48:24 Main [App] >> Version: 7.2.3
(2) Sep07 20:48:24 Main [App] >> LocalNode: beegfs-mgmtd ctrl0 [ID: 1]
(2) Sep07 20:48:24 Main [App] >> Usable NICs: ens2(TCP)
metadata0 $ cat /var/log/beegfs-meta.log
(3) Sep07 20:45:12 Main [App.cpp:641] >> Limiting number of xattrs per inode.
(3) Sep07 20:45:12 Main [App] >> Root directory loaded.
(3) Sep07 20:45:12 Main [RegDGramLis] >> Listening for UDP datagrams: Port 8005
(1) Sep07 20:45:12 Main [App] >> Waiting for beegfs-mgmtd@ctrl0:8008...
I can see udp packets arriving on port 8008 of the management node when the beegfs-meta service start.
When I install and start beegfs-meta on the same node as mgmtd, the two services communicate successfully.
Any ideas or suggestions for troubleshooting?
Best regards,
Nicolas