Unable to start a One Master Server and One Volume Server: port 19333:connection refused

Question

I tried to build a simple one master server and one volume server according to the "Quick Start with Single Binary" section in official README.md, but failed on port 19333 connection refused.

Environment

I tried both on a local VMWare machine and an Aliyun ECS server. Their system version are as the following:

Local VM

Linux version 5.15.0-67-generic (buildd@lcy02-amd64-116) (gcc (Ubuntu 11.3.0-1ubuntu1~22.04) 11.3.0, GNU ld (GNU Binutils for Ubuntu) 2.38) #74-Ubuntu SMP Wed Feb 22 14:14:39 UTC 2023

Aliyun ECS

Linux version 4.15.0-206-generic (buildd@lcy02-amd64-070) (gcc version 7.5.0 (Ubuntu 7.5.0-3ubuntu1~18.04)) #217-Ubuntu SMP Fri Feb 3 19:10:13 UTC 2023

Install Seaweedfs

Download version 3.44 linux_amd64.tar.gz from the release page.

$ tar -xvf linux_amd64.tar.gz
$ chown root:root weed
$ mv weed /usr/local/bin
$ ls -lah /usr/local/bin/
total 79M
drwxr-xr-x  2 root root 4.0K Mar 31 12:19 .
drwxr-xr-x 10 root root 4.0K Apr 21  2022 ..
-rwxr-xr-x  1 root root  79M Mar 20 03:41 weed
$ weed version
version 30GB 3.44 5b43c4bb98c5e2d83dca5414b4c902b668f91ea5 linux amd64

Run server

Run weed server -dir=./data -s3 get the following error:

$ mkdir data
$ weed server -dir=./data -s3
I0331 13:45:28.688352 master.go:269 current: 192.168.0.92:9333 peers:
I0331 13:45:28.689331 file_util.go:23 Folder ./data Permission: -rwxrwxr-x
I0331 13:45:28.689812 file_util.go:23 Folder ./data Permission: -rwxrwxr-x
I0331 13:45:28.690026 master.go:269 current: 192.168.0.92:9333 peers:192.168.0.92:9333
I0331 13:45:28.690501 master_server.go:127 Volume Size Limit is 30000 MB
I0331 13:45:28.691132 volume_grpc_client_to_master.go:43 checkWithMaster 192.168.0.92:9333: get master 192.168.0.92:9333 configuration: rpc error: code = Unavailable desc = connection error: desc = "transport: Error while dialing: dial tcp 192.168.0.92:19333: connect: connection refused"
I0331 13:45:28.691515 master.go:150 Start Seaweed Master 30GB 3.44 5b43c4bb98c5e2d83dca5414b4c902b668f91ea5 at 192.168.0.92:9333
I0331 13:45:28.692629 raft_server.go:118 Starting RaftServer with 192.168.0.92:9333
I0331 13:45:28.694544 raft_server.go:167 current cluster leader:
I0331 13:45:30.482563 volume_grpc_client_to_master.go:43 checkWithMaster 192.168.0.92:9333: get master 192.168.0.92:9333 configuration: rpc error: code = Unavailable desc = connection error: desc = "transport: Error while dialing: dial tcp 192.168.0.92:19333: connect: connection refused"
I0331 13:45:30.690445 s3.go:177 wait to connect to filer 192.168.0.92:8888 grpc address 192.168.0.92:18888
I0331 13:45:31.692348 s3.go:177 wait to connect to filer 192.168.0.92:8888 grpc address 192.168.0.92:18888
I0331 13:45:32.273947 volume_grpc_client_to_master.go:43 checkWithMaster 192.168.0.92:9333: get master 192.168.0.92:9333 configuration: rpc error: code = Unavailable desc = connection error: desc = "transport: Error while dialing: dial tcp 192.168.0.92:19333: connect: connection refused"

I hope the server could start properly.

Firewall has been turned off and iptable is clean, so I think no port is blocked.

$ sudo iptables -nL
Chain INPUT (policy ACCEPT)
target     prot opt source               destination

Chain FORWARD (policy ACCEPT)
target     prot opt source               destination

Chain OUTPUT (policy ACCEPT)
target     prot opt source               destination
$ sudo ufw status
Status: inactive

Answer 1

I once deleted the post, but consider for a while and make it visible again.

Just want to make anyone new to seaweedfs to know, it is normal to have grpc connection error when you just start the server. Please wait for a 1-2 minutes to see whether the system works properly.

--- Want to sharing another case when I meet when started up the my service with multiple nodes ---

If you have a cluster with multiple masters, when start a separate volume server or filer server to connect to all of them as

weed filer -s3 -ip=192.168.9.16 -port=8888 -master=192.168.9.16:9333,192.168.9.17:9333,192.168.9.18:9333

The api http://192.168.9.16:8888 will be available only after all three masters has been started.

This restriction only happens when starting up the filer server. If later the master leader failover to another master node, the filer server will not be affected.