☰

从 squeeze 升级到 jessie 后出现网卡错误

tag-icon • tag-icon linux debian network-adapter
从 squeeze 升级到 jessie 后出现网卡错误

我们将一个非常过时的中央路由器从apt-get dist-upgradeDebian Squeeze over Wheezy 升级到了 Jessie。

但现在我们看到以下错误dmesg：

[Tue Apr 12 22:17:55 2016] e1000 0000:09:03.0 wi22: Detected Tx Unit Hang
  Tx Queue             <0>
  TDH                  <2>
  TDT                  <5b>
  next_to_use          <5b>
  next_to_clean        <0>
buffer_info[next_to_clean]
  time_stamp           <3abbd6>
  next_to_watch        <4>
  jiffies              <3abd34>
  next_to_watch.status <0>
[Tue Apr 12 22:17:57 2016] e1000 0000:09:03.0 wi22: Detected Tx Unit Hang
  Tx Queue             <0>
  TDH                  <2>
  TDT                  <5b>
  next_to_use          <5b>
  next_to_clean        <0>
buffer_info[next_to_clean]
  time_stamp           <3abbd6>
  next_to_watch        <4>
  jiffies              <3abf29>
  next_to_watch.status <0>
[Tue Apr 12 22:17:59 2016] e1000 0000:09:03.0 wi22: Detected Tx Unit Hang
  Tx Queue             <0>
  TDH                  <2>
  TDT                  <5b>
  next_to_use          <5b>
  next_to_clean        <0>
buffer_info[next_to_clean]
  time_stamp           <3abbd6>
  next_to_watch        <4>
  jiffies              <3ac11e>
  next_to_watch.status <0>
[Tue Apr 12 22:18:01 2016] e1000 0000:09:03.0 wi22: Detected Tx Unit Hang
  Tx Queue             <0>
  TDH                  <2>
  TDT                  <5b>
  next_to_use          <5b>
  next_to_clean        <0>
buffer_info[next_to_clean]
  time_stamp           <3abbd6>
  next_to_watch        <4>
  jiffies              <3ac313>
  next_to_watch.status <0>
[Tue Apr 12 22:18:03 2016] e1000 0000:09:03.0 wi22: Reset adapter
[Tue Apr 12 22:18:03 2016] br50: port 2(wi22.50) entered disabled state
[Tue Apr 12 22:18:03 2016] br60: port 2(wi22.60) entered disabled state
[Tue Apr 12 22:18:03 2016] br161: port 1(wi22.161) entered disabled state
[Tue Apr 12 22:18:03 2016] br170: port 7(wi22.170) entered disabled state
[Tue Apr 12 22:18:03 2016] e1000: wi22 NIC Link is Up 1000 Mbps Full Duplex, Flow Control: None
[Tue Apr 12 22:18:03 2016] br50: port 2(wi22.50) entered forwarding state
[Tue Apr 12 22:18:03 2016] br50: port 2(wi22.50) entered forwarding state
[Tue Apr 12 22:18:03 2016] br60: port 2(wi22.60) entered forwarding state
[Tue Apr 12 22:18:03 2016] br60: port 2(wi22.60) entered forwarding state
[Tue Apr 12 22:18:03 2016] br161: port 1(wi22.161) entered forwarding state
[Tue Apr 12 22:18:03 2016] br161: port 1(wi22.161) entered forwarding state
[Tue Apr 12 22:18:03 2016] br170: port 7(wi22.170) entered forwarding state
[Tue Apr 12 22:18:03 2016] br170: port 7(wi22.170) entered forwarding state

所有 1Gbit 光纤接口大约每 2 秒就会发生一次这种情况。

一些用户甚至报告了连接不稳定。

这是什么意思？是因为硬件过时了吗？

我们打算买一个新的，但是在新路由器出来之前怎么解决这个问题呢？

答案1

错误的原因可能是当前内核中的一个错误。

https://sourceforge.net/p/e1000/bugs/432/

我们将内核降级为3.2.0-4并且它再次可以运行。

相关内容