[
https://issues.apache.org/jira/browse/SSHD-711?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Goldstein Lyor resolved SSHD-711.
---------------------------------
Resolution: Not A Problem
> Apache hangs sending file
> -------------------------
>
> Key: SSHD-711
> URL: https://issues.apache.org/jira/browse/SSHD-711
> Project: MINA SSHD
> Issue Type: Bug
> Affects Versions: 1.1.0
> Reporter: Roger Hart
> Priority: Minor
> Attachments: SSHDbad.log, SSHDgood.log
>
>
> We're running 1.1.0 but I get the same problem with 1.2.0. When sending a
> file to a particular server we have problems on some machines but not others.
> I've attached a DEBUG log for a "good" machine and a "bad" machine. For the
> "bad" machine, the system hangs midway through sending the file. Here are the
> final log entries.
> My theory is that "normally" you see a "Writing X bytes" message, followed by
> a "Finished writing" message, followed by a "Read Y bytes" message. But on
> the "bad" machine there seems to be 2 "Write" messages before the hang. I
> think this is because there are 2 threads putting a write request in the
> write queue and both are written out to the socket. I'm wondering if this can
> cause the max packet size specified in the SSH_MSG_CHANNEL_OPEN_CONFIRMATION
> message to be exceeded. The "good" machine has fewer threads than the "bad"
> machine, and we've also observed that a smaller files is less likely to hit
> the problem.
> Do you think this theory is valid?
> Thanks,
> Roger
> 10:29:22.448 7-Nov-16 -0800 | (lient.channel.ChannelSubsystem)
> handleData(ChannelSubsystem[id=0,
> recipient=0]-ClientSessionImpl[USER@/X.X.X.X:22][sftp]) SSH_MSG_CHANNEL_DATA
> len=19
> 10:29:22.450 7-Nov-16 -0800 | (che.sshd.common.channel.Window)
> waitForSpace(ChannelSubsystem[id=0,
> recipient=0]-ClientSessionImpl[USER@/X.X.X.X:22][sftp]: client remote window)
> available: 1073741779
> 10:29:22.450 7-Nov-16 -0800 | (che.sshd.common.channel.Window)
> waitAndConsume(ChannelSubsystem[id=0,
> recipient=0]-ClientSessionImpl[USER@/X.X.X.X:22][sftp]: client remote window)
> - requested=32768, available=1073741779
> 10:29:22.463 7-Nov-16 -0800 | (shd.common.io.nio2.Nio2Session) Writing 32820
> bytes
> 10:29:22.464 7-Nov-16 -0800 | (che.sshd.common.channel.Window)
> waitForSpace(ChannelSubsystem[id=0,
> recipient=0]-ClientSessionImpl[USER@/X.X.X.X:22][sftp]: client remote window)
> available: 1073709011
> 10:29:22.464 7-Nov-16 -0800 | (che.sshd.common.channel.Window)
> waitAndConsume(ChannelSubsystem[id=0,
> recipient=0]-ClientSessionImpl[USER@/X.X.X.X:22][sftp]: client remote window)
> - requested=31, available=1073709011
> 10:29:22.464 7-Nov-16 -0800 | (shd.common.io.nio2.Nio2Session) Writing 84
> bytes
> 10:29:22.534 7-Nov-16 -0800 | (shd.common.io.nio2.Nio2Session) Finished
> writing
> 10:29:22.534 7-Nov-16 -0800 | (shd.common.io.nio2.Nio2Session) Finished
> writing
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
