Sometimes, snmpd will report the wrong speed of a network interface. For example, the following output shows a speed of 10Mbps, whereas the ethernet port on the server is actually working at 100Mbps:

    IF-MIB::ifIndex.2 = INTEGER: 2
    IF-MIB::ifDescr.2 = STRING: eth0
    IF-MIB::ifType.2 = INTEGER: ethernetCsmacd(6)
    IF-MIB::ifMtu.2 = INTEGER: 1500
    IF-MIB::ifSpeed.2 = Gauge32: 10000000

The Net-SNMP documentation mentions this. This can be manually fixed with the interface directive in snmpd.conf, for example:

    interface eth0 6 100000000

Notes:

• lb_algo determines which real server gets to serve a request, and possible algorithms include round-robin (weighted or not), least-connections (weighted or not), etc (see RHEL docs).
• lb_kind determines how the real server receives the request, and the supported methods on RHEL 7 are NAT and DR.

lb_kind DR

Notes:

• The load balancers and the real servers need to be on the same network segment
• The Virtual IP is configured on all real servers, and real servers need to be configured to ignore ARP requests for the virtual IP. This can be achieved with either arptables, or iptables or sysctl configuration.

See also

• keepalived documentation for RHEL 7

This article has helped me create bootable Windows installation USBs when installating from an optical drive was not an option:

• Create A Bootable Windows 7 USB Drive From Linux (Tested On Ubuntu)

CentOS 6 is not directly supported by version 4 of the VMware ESXi hypervisor, which means that there are no precompiled binaries for the kernel modules required for the VM to talk to the host. This short guide shows you how to manually compile the tools. Note that every time that you upgrade to a newer kernel version, you will have to do this from scratch -- you will have to compile the tools against the new kernel's headers.

Here's the procedure:

1. From the vSphere client, right click on the machine, open the "Guest" menu and select "Install/Upgrade VMware Tools". This will insert a virtual optical drive in the VM.

2. Mount the virtual optical drive, not directly on the /mnt directory, since that will be used later by the installer. For example:

   mkdir /mnt/vmware-tools
   mount -o loop /dev/cdrom /mnt/vmware-tools
   

3. Copy the tools in a writeable location (the mount point of the virtual optical drive will be read-only) and decompress it. For example:

   cp /mnt/vmware-tools/VMwareTools-4*.tar.gz /root/
   cd /root/
   tar zxvf VMwareTools-4*.tar.gz
   

4. Before starting the installation, install a C compiler and the kernel's source code, since the VMware tools installer will need the kernel headers to include while compiling the modules. For example:

   yum install gcc
   yum install kernel-headers
   

5. Next, start the installation. It will ask a lot of questions, on which you can (most probably) accept the default answers.

   cd /root/vmware-tools-distrib
   ./vmware-install.pl
   

6. Finally, even though you don't normally need to reboot, do it anyway if you can. That way you will verify that the kernel modules are properly loaded after a reboot. You can check that with:

   [[email protected] ~]# lsmod | grep '^vm'
   vmmemctl                8642  0 
   vmware_balloon          7199  0
   

7. If everything went fine, you might want to clean up the files you copied and extracted, although keeping them will speed up the reinstallation of the tools in case you upgrade the kernel to a newer version.

As a note to myself and a future reference, here is a list of features of VMware ESXi 6 that get disabled when the 60-day evaluation period expires.

• True Zero Downtime HAProxy Reloads
• True Zero Downtime HAProxy Reloads: An Alternative Approach

Step by step guides on how to install Zimbra on either CentOS 7 or Ubuntu.

http://sysadvent.blogspot.com/2014/12/day-3-so-server-tell-me-about-yourself.html

• Tutorial to create a vector graphic with Inkscape

After physically replacing the disk, the new disk might not be assigned to the controller on which the old disk was assigned, depending on the value of the disk.auto_assign option, which you can check with:

options disk.auto_assign

Even if the value of that option is on, the disk might still remain unassigned, in which case you will see a message for unassigned disks in the end of the output of the command:

disk show

You can see which disks are unassinged with

disk show -n

To assign a disk to a controller, SSH to that controller and do:

disk assign XX.YY.ZZ

... where XX.YY.ZZ is the name of the disk, as obtained by disk show -n. Example output:

FAS> disk assign 01.23.45
Fri May 13 00:00:02 [FAS:diskown.changingOwner:info]: changing ownership for disk 12.34.45 (S/N ABCDEF) from unowned (ID 1234567890) to FAS (ID 0987654321)