Heres a difficult one =]
I've got an Win2k SP2 box uploading and downloading via SMB sessions to a Linux RedHat 6. 1 RAID array (running Samba) some pretty meaty files, from 30MB to +400MB. The network is on a complete (with zero to moderate traffic at the time of the tests) 100Mbit switch with both machines negotiating a 100/Full connection.
I've got an Win2k SP2 box uploading and downloading via SMB sessions to a Linux RedHat 6.1 RAID array (running Samba) some pretty meaty files, from 30MB to +400MB. The network is on a complete (with zero to moderate traffic at the time of the tests) 100Mbit switch with both machines negotiating a 100/Full connection.
The issue here is the uploads top out at 150KB/sec (for both an SMB (Picture) session and an FTP (Picture) session) and CPU utilization will be ~99.9% on the Linux side. If a file is being downloaded it will generally reach a mere 3MB/sec, which is acceptable compared to the uploads. If an upload and a download is taking place, the download will ride within the transfer thresh hold of the upload (as monitored from the Win2k box). (Picture)
This appears to be not so much an NT question as it is just a general question on TCP/IP logic..
The issue here is the uploads top out at 150KB/sec (for both an SMB (Picture) session and an FTP (Picture) session) and CPU utilization will be ~99.9% on the Linux side. If a file is being downloaded it will generally reach a mere 3MB/sec, which is acceptable compared to the uploads. If an upload and a download is taking place, the download will ride within the transfer thresh hold of the upload (as monitored from the Win2k box). (Picture)
This appears to be not so much an NT question as it is just a general question on TCP/IP logic..
Participate on our website and join the conversation
This topic is archived. New comments cannot be posted and votes cannot be cast.
Responses to this topic
I can verify that the user with upload and download privilages has no bandwidth restrictions..
I have considered it being some sort of server-side parity check, however, it taking place with just the uploads and then having a negative effect on the downloads when both sessions are active disuades me from that conclusion.
I have considered it being some sort of server-side parity check, however, it taking place with just the uploads and then having a negative effect on the downloads when both sessions are active disuades me from that conclusion.
Have you tryed messing about your network settings? is it properly configured.
I remember seeing/hearing this on another occasion. However there was no solusion as far as i can recall.
I do not use Linux except 3-4 Cobalt Racks and we do not have this problem. But we have sh1t loads of others.
If you ask me Linux just does not cut it as a serious server SW. Why bother with it when theer is Unix and W2k
I remember seeing/hearing this on another occasion. However there was no solusion as far as i can recall.
I do not use Linux except 3-4 Cobalt Racks and we do not have this problem. But we have sh1t loads of others.
If you ask me Linux just does not cut it as a serious server SW. Why bother with it when theer is Unix and W2k
