Change History (12)
comment:1 by , 11 years ago
| Owner: | changed from to |
|---|---|
| Status: | new → assigned |
comment:2 by , 11 years ago
Doesn't work. Statd fails to start, which means that locking doesn't work. My fstab specifies locking, so the mounts fail. daemon.log has messages like:
rpc.statd[nnnn]: failed to create rpc listeners, exiting
where nnnn is the pid. I have four of those from the initial failure when starting and the three mount failures.
comment:3 by , 11 years ago
The rpcbind bootscript reports success, but it isn't running. The same thing if I run rpcbind manually - rc 0 but no process. The nfs-client bootscript then reports success, but fails.
On shutdown, both statd and rpcbind are reported as shutting down correctly.
follow-up: 5 comment:4 by , 11 years ago
Does the previous version still work in the new environment?
follow-up: 6 comment:5 by , 11 years ago
Replying to bdubbs@…:
Does the previous version still work in the new environment?
Yes, I'm using 0.2.0 on my first 7.4 builds.
comment:6 by , 11 years ago
Replying to ken@…:
Replying to bdubbs@…:
Does the previous version still work in the new environment?
Yes, I'm using 0.2.0 on my first 7.4 builds.
But I've just built 0.2.0 on the new system, to get an idea how the trace ought to look, and that one doesn't work either. I suppose it has to be a change in the kernel headers between 3.10.7 and 3.10.9.
comment:8 by , 11 years ago
The daemon log said something like:
Aug 26 13:00:54 blfs rpcbind: cannot get uid of '': Success
I took a look at the source :) and it appears that there is something wrong with the code that assumes the root user. Try configuring with --with-rpcuser=root. That seems to get it to run for me. I haven't done any additional testing, but that may give you some help.
Note that I did run with: sudo /sbin/rpcbind -d and got a couple of messages:
can't get local ip4 address: Servname not supported for ai_socktype can't get local ip6 address: Servname not supported for ai_socktype
I'll look at that next.
comment:9 by , 11 years ago
I see that the "can't get local ip? address" message is caused by the instruction:
getaddrinfo(NULL, "sunrpc", &hints, &res)
We removed sunrpc from services, so I think the message is not relevant.
Builds with current instructions. I'll give it a try in my next build (probably, rebuilding current LFS from -rc1). That may be a few days away.