Tweet
Diskless + suspend to ram = 4 min resume time
Hi everyone, I've a question for Phoronix experts. I've a machine that boots via PXE (nfs root) and then stays in S3 (suspend to ram) most of the time. The problem is, resuming often takes up to 4 minutes, far more than a cold boot. I said often, because sometimes it takes just 10 seconds, as it should be. Server load is at about 10%, the connection is a gigabit one. I noticed that the resume is usually faster if the client has been suspended only for a few seconds.
Relevant logs:
CLIENT dmesg
Mar 3 17:25:35 mce kernel: [17643.650102] nfs: server 192.168.1.1 not responding, still trying
Mar 3 17:29:35 mce kernel: [17883.652715] nfs: server 192.168.1.1 OK
SERVER dmesg
[9075279.281091] nfsd: last server has exited, flushing export cache
[9075280.699313] svc: failed to register lockdv1 RPC service (errno 97).
[9075280.700525] NFSD: Using /var/lib/nfs/v4recovery as the NFSv4 state recovery directory
[9075280.715802] NFSD: starting 90-second grace period
I've tried changing the timeout on the client (from 5 seconds to 120) but it didn't help.
Thanks in advance for any help, it's not good to wait 4 minutes just to see the news...
Relevant logs:
CLIENT dmesg
Mar 3 17:25:35 mce kernel: [17643.650102] nfs: server 192.168.1.1 not responding, still trying
Mar 3 17:29:35 mce kernel: [17883.652715] nfs: server 192.168.1.1 OK
SERVER dmesg
[9075279.281091] nfsd: last server has exited, flushing export cache
[9075280.699313] svc: failed to register lockdv1 RPC service (errno 97).
[9075280.700525] NFSD: Using /var/lib/nfs/v4recovery as the NFSv4 state recovery directory
[9075280.715802] NFSD: starting 90-second grace period
I've tried changing the timeout on the client (from 5 seconds to 120) but it didn't help.
Thanks in advance for any help, it's not good to wait 4 minutes just to see the news...
Comment