Home / user-298003

EM0's questions

Martin Hope
EM0
Asked: 2020-03-12 08:22:10 +0800 CST
  • 2

In Ubuntu 18.04 I have the netplan bonding configuration like this:

  bonds:
    bond0:
      interfaces:
        - eno1
        - eno2
      parameters:
        mode: active-backup
        primary: eno1

I discovered that this does not actually cause the network to fail over to eno2 if the cable is unplugged from eno1 unless I set mii-monitor-interval, like this:

  bonds:
    bond0:
      interfaces:
        - eno1
        - eno2
      parameters:
        mode: active-backup
        mii-monitor-interval: 100

(Setting primary seems to be unnecessary - is it?)

However, this change does not seem to take effect after netplan apply - cat /proc/net/bonding/bond0 does not show the new MII Polling Interval value until I reboot.

Is there a way to apply this change without rebooting?

https://unix.stackexchange.com/questions/148838/reset-ethernet-bonding-without-rebooting suggests service network restart but there is no "network" service in Ubuntu 18.04.

networking network-bonding
Martin Hope
EM0
Asked: 2015-03-13 08:50:21 +0800 CST
  • 2

I have Ubuntu 12.04 (x86) running under VMware Workstation 10.0.5 on a Windows 7 x64 host. Disk access in this VM is noticeably slow - it takes ages just to start Firefox, for example, and the disk light is on the whole time. hdparm -tT confirms this:

/dev/sda:
 Timing cached reads:   18100 MB in  1.99 seconds = 9073.75 MB/sec
 Timing buffered disk reads:  42 MB in  3.54 seconds =  11.85 MB/sec

However, I have 2 other VMs on the same host partition (Ubuntu 11.04 and 14.04) and they're both OK:

/dev/sda:
 Timing cached reads:   17388 MB in  2.00 seconds = 8703.76 MB/sec
 Timing buffered disk reads: 210 MB in  3.07 seconds =  68.37 MB/sec

/dev/sda:
 Timing cached reads:   17012 MB in  2.00 seconds = 8511.87 MB/sec
 Timing buffered disk reads: 282 MB in  3.01 seconds =  93.63 MB/sec

The big difference is that the slow machine has been used a lot over time and had a lot of stuff installed, so the virtual disk is 32GB (vs 4GB and 6GB for the other VMs). What can I do to restore disk performance? I tried the VMware De-fragment and Compact commands that they didn't help. The file system is ext4 in all 3 cases.

performance