Hello,

 

Recently my unraid server has started acting up when streaming media.  Intermittently the media will freeze and my user share's will briefly drop off my network.

 

In the syslogs, I found the following corresponding entries:

 

Nov  6 21:20:38 jNAS kernel: e1000: eth0: e1000_clean_tx_irq: Detected Tx Unit Hang

Nov  6 21:20:38 jNAS kernel:  Tx Queue            <0>

Nov  6 21:20:38 jNAS kernel:  TDH                  <43>

Nov  6 21:20:38 jNAS kernel:  TDT                  <47>

Nov  6 21:20:38 jNAS kernel:  next_to_use          <47>

Nov  6 21:20:38 jNAS kernel:  next_to_clean        <40>

Nov  6 21:20:38 jNAS kernel: buffer_info[next_to_clean]

Nov  6 21:20:38 jNAS kernel:  time_stamp          <46f85>

Nov  6 21:20:38 jNAS kernel:  next_to_watch        <43>

Nov  6 21:20:38 jNAS kernel:  jiffies              <47245>

Nov  6 21:20:38 jNAS kernel:  next_to_watch.status <0>

Nov  6 21:20:39 jNAS ifplugd(eth0)[1314]: Link beat lost.

Nov  6 21:20:42 jNAS kernel: e1000: eth0 NIC Link is Up 1000 Mbps Full Duplex, Flow Control: RX/TX

Nov  6 21:20:43 jNAS ifplugd(eth0)[1314]: Link beat detected.

 

 

Doing some research online, I found the following thread among a bunch of not as useful suggestions:

 

http://blog.bradiceanu.net/2010/11/28/netdev-watchdog-eth0-transmit-timed-out/

 

What caught my eye here is that these problems just started occurring recently and, in fact, coincide with me upgrading the machine to 4GBs of RAM. (Which that thread identifies as a potential problem point.)

 

The suggested solution says that I will require Intel’s e1000 driver version 8.0.25 or higher.

 

I am running unRaid version 4.7 and my driver version is as follows:

 

jNAS:/var/log# ethtool -i eth0

driver: e1000

version: 7.3.21-k5-NAPI

 

Is it possible to upgrade this driver, or am I stuck either removing some RAM or finding a new NIC?

 

Thanks.

unRAID version 5.0 uses the newer e1000e driver.

Fair enough.  Thanks for the info.