cancel
Showing results for 
Search instead for 
Did you mean: 

CIFS limitation?

greglenain
Champ in-the-making
Champ in-the-making
When I try to copy a set of files (49 docs, 15 Mo) through CIFS, Alfresco closes the port and raises the following exception:

Winsock rx2 len=49
10:24:32,210 ERROR [alfresco.smb.protocol] Closing session due to exception
org.alfresco.filesys.netbios.win32.WinsockNetBIOSException: null - Message too long
  at org.alfresco.filesys.smb.server.win32.WinsockNetBIOSPacketHandler.readPacket(WinsockNetBIOSPacketHandler.java:138)
        at org.alfresco.filesys.smb.server.SMBSrvSession.runSMBSrvSession.java:1511)
        at java.lang.Thread.run(Thread.java:595)

Only 20% of the files are processed before the problem occurs.

I operate that copy from a Window OS to an Alfresco 1.4 repository.

Are there some limitations in the CIFS usage ?

Thanks.
16 REPLIES 16

gary_spencer
Champ in-the-making
Champ in-the-making
Hi,

Could you give us some more details on your OS version/service pack and network setup.

Cheers

Gary

greglenain
Champ in-the-making
Champ in-the-making
Client side:
    OS: Windows XP, SP2
    IP conf: DHCP
Repo side:
    Alfresco 1.4 CE
    OS: Windows 2000, SP4
    HTTP Srv: Apache Tomcat 5.5.12
    Java: Sun 1.5.0 05
    IP conf: DHCP
Network:
    LAN: 100Mb
    Domain server: Linux Red Hat 4
Furthermore, the copy process succeeds when using FTP instead of CIFS.

greglenain
Champ in-the-making
Champ in-the-making
No news is good news ?

rdanner
Champ in-the-making
Champ in-the-making
When I try to copy a set of files (49 docs, 15 Mo) through CIFS, Alfresco closes the port and raises the following exception:

Winsock rx2 len=49
10:24:32,210 ERROR [alfresco.smb.protocol] Closing session due to exception
org.alfresco.filesys.netbios.win32.WinsockNetBIOSException: null - Message too long
  at org.alfresco.filesys.smb.server.win32.WinsockNetBIOSPacketHandler.readPacket(WinsockNetBIOSPacketHandler.java:138)
        at org.alfresco.filesys.smb.server.SMBSrvSession.runSMBSrvSession.java:1511)
        at java.lang.Thread.run(Thread.java:595)

Only 20% of the files are processed before the problem occurs.

I operate that copy from a Window OS to an Alfresco 1.4 repository.

Are there some limitations in the CIFS usage ?

Thanks.

you are picking up 19 docs and dropping them in a cifs projection? can you identify that its the number of docs or one of the docs in particular?

greglenain
Champ in-the-making
Champ in-the-making
In that example, I've picked up 49 docs from a local filesystem and dropped them in a CIFS directory bound to my client files explorer.
And "only 20% of the files were processed".

Neither 49 nor 20% are magic numbers.

Sometimes, Alfresco may copy up to 30%, maybe 40% of the files.
But other times, the repo can't climb further than 15%.
And I can see no huge (or corrupted or evil) file hidden among the others.

Let's sum up: opposite to FTP, CIFS seems not to be able to stand the load.

Can users rely on CIFS ?

rdanner
Champ in-the-making
Champ in-the-making
In that example, I've picked up 49 docs from a local filesystem and dropped them in a CIFS directory bound to my client files explorer.
And "only 20% of the files were processed".

Neither 49 nor 20% are magic numbers.

Sometimes, Alfresco may copy up to 30%, maybe 40% of the files.
But other times, the repo can't climb further than 15%.
And I can see no huge (or corrupted or evil) file hidden among the others.

Let's sum up: opposite to FTP, CIFS seems not to be able to stand the load.

Can users rely on CIFS ?

The reliability of this interface is hugely important to me.  Is there a response on this?

btorfs
Champ in-the-making
Champ in-the-making
Hi;
We have the same message in our alfresco logfiles when cifs crashes.
Is there any progress on this?

esanamr
Champ in-the-making
Champ in-the-making
We have the same problem.

If you solve it, please tell us.

Thanks in advance

gary_spencer
Champ in-the-making
Champ in-the-making
Hi,

We're trying to get hold of a packet trace from a site that has this problem as we're not able to reproduce the problem.

If you want to send an Ethereal packet trace to me it might help speed up sorting out a fix.

Cheers

Gary