vCSA default shell is BASH


A quick post about a little caveat while working in the vCenter Server Appliance (vCSA) shells. Yes correctly, shells in plural. The vCSA is bundled with at least two different shells:

  • Appliance Shell (default)
  • BASH shell

The “Appliance shell” is the default shell. After you log in to the vCSA, it will present the following well known screen.
Fig 1.

The appliance shell can be used for updating the vCSA, using the software-packages command and has some other use cases. From here you can enable the BASH shell as shown in the Fig 1. for the duration of your session with the following commands:

# shell.set --en -s /bin/bash root
# chsh -s /bin/bash root

You can also set the BASH shell as the default shell by performing the following command. Make sure, you first enable the BASH shell as shown above:

# chsh -s /bin/bash root

For the change to take effect, log out and log in again. Now you will directly enter the BASH shell.

But while working in the BASH shell, you need to temporarily switch to the Appliance shell?
In that case, provide the following command:

# appliancesh

That’s it. A shell is nothing more or less than an executable; the “Appliance shell” is no exception and can be found as /bin/appliancesh.

For more information, see: VMware KB “Toggling the vCenter Server Appliance 6.x default shell (2100508)

Best way to create vCSA support bundles


In general, during contact with a Customer Support team, whether being a Hardware vendor (Servers, Storage) or a Software vendor, the likelihood that you will be asked to upload some log files for further investigation is significantly.

In case of VMware vSphere, you will have multiple options for collecting a support bundle; to name a few:

  • Windows vSphere Client
  • vSphere Web Client
  • PowerShell scripts

See also VMware KB “Collecting diagnostic information for VMware vCenter Server 4.x, 5.x and 6.x (1011641)” for a complete overview.

Using some of these methods is very convenient; however there is a little caveat, when the vCenter Server is a vCenter Server Appliance (vCSA) 6.x.

Under certain conditions, a vCSA might contain core dump files. When requested to create a support bundle these core dump files will be added to the support bundle, together with the log files. The issue that may arise is that the location where the support bundle will be created (partition /storage/log) has a fixed size and possibly is too small.

Is that is the case, the creation of the support bundle will halt with error “Cannot create a diagnostic bundle” and the desired support bundle will not be created.

VMware recommendation is to create and download a support bundle using the web browser. To do so enter the following URL:
https ://<VCSA Hostname or IP address>/appliance/support-bundle

Fig 1

After providing the credentials, the support-bundle (filename is: vm-support.tgz) will start downloading. The progress of the process will be shown in the browser.

Using this method, the files will be directly downloaded to your local computer, instead of being prepared on the vCSA.

As always, I thank you for reading.

All Hands on Deck / VMware


201701-01Although VMware is best known for its core virtualization product vSphere, it may not come as a surprise that since their first product releases, a lot has changed. Over the years, on top of this solid foundation, VMware built a “Software Defined Everything” stack (SDDC), a Cloud stack with all the Automation and Orchestration tools, Monitoring tools, an End User Compute environment etc. This has led to a very large number of VMware products.
As you are as curious as I am, product
data sheets and documentation maybe helpful, but the best way in my humble opinion is to download, install and try the bytes yourself.
In addition to options already there, VMware launched a new one; Product Walkthroughs. So time for an overview how to get familiar with these fine products.

Read the rest of this entry »

Windows 10 dual-boot issues


Usually my writings are VMware related, so the title of this post seams to be off-topic. However VMware products played an important role in solving my dual-boot issues.
I am not going into great detail about the “Why do I want to install another OS on my computer?”. Well, I am not fully satisfied with Windows 10, but I have not reached the point to completely wipe Windows 10 off my Computer (an Asus N56VM with the original HDD replaced by a SSD) and do a fresh Linux install. For that reason, I decided to share my disk with the latest Ubuntu 16.04 LTS version and have both OS available.

In a nutshell, the first steps are:

  • Free up a block of continuous disk space to make room for the Ubuntu installation.
  • Download Ubuntu and create a bootable USB drive or DVD.
  • Install Ubuntu in the available disk space.

During the installation of Ubuntu, you are presented the option to install Ubuntu alongside your existing Windows installation. The installation went smoothly, but after rebooting my computer I found out that Windows 10 was the one and only available bootable OS.

The Internet presents an overwhelming amount of advice for solving dual-boot issues. It is important to know if your computer has a BIOS or a UEFI firmware, in my case UEFI it is.
A very common advice is to switch-off Secure Boot, but the Asus UEFI firmware does not have this option. Also, major Linux distributions do support Secure Boot, so this does not make sense imho. See this link for more information about UEFI.

Read the rest of this entry »

Check_MK and vSphere – vCenter Server


This post is the third part in a series about Check_MK and vSphere. In the second part, I showed you the options for monitoring an ESXi host without using vCenter Server. In this post we will explore the options for monitoring a vCenter Server on Windows and also the vCenter Server Appliance (VCSA).

vCenter Server Windows

For this POC we have a vCenter Server 6.0 U2 (build 3634793) on a Windows Server 2012 R2. As this is a normal Windows server, I installed the Check_MK agent for Windows and added the host to Check_MK. For the property Agent type, select “Check_MK Agent (Server)”.

2016-08-21_01Figure 1

By default the Check_MK Windows Agents presents – without further tweaking – a lot of information; CPU and Memory utilization, Disk I/O, status of the filesystems, status of the Network interfaces and many more.

Now it’s time to reveal the vSphere options. We follow the same procedure as we did for the ESXi host. In the WATO configuration go to Host & Service Parameters \ Datasource Programs and select Check state of VMware ESX via vSphere. Now create a second rule for the vCenter Server, start with providing a descriptive name.

Read the rest of this entry »

Check_MK and vSphere – ESXi


This post is the second part in a series about Check_MK monitor and vSphere. In the first part Check_MK was introduced and some basic Installation and Configuration was explained.

According to the documentation, for monitoring VMware ESXi and vCenter Server, Check_MK has implemented a plugin that uses the vSphere API for retrieving information. This plugin is much more efficient than versions based on the Perl API.

So let’s start and see what can be revealed. To get a clear understanding of the various options, I will perform a step-by-step configuration instead of ticking all options at once.

The first step is to go into WATO and add an ESXi host. Under WATO, choose, Hosts and New Host.

2016-08-10-01Figure 1

You must at least enter the Hostname and an IP address, the Alias is optional. Under Agent Type place a tick and select “No Agent”.

At this time, the result is not very exciting; the ESXi host will be pinged.

2016-08-10-02Figure 2

Read the rest of this entry »

Check_MK and vSphere – Introduction



Infrastructure monitoring is essential to carry out proper System Administration. Infrastructure consists of many components, starting with the basics such as server hardware, network components, storage, uninterruptible power supplies, backup equipment, but also environmental factors such as temperature and air humidity in server rooms. Apart from understanding the hardware, software is the next layer. Starting with Operating Systems; monitoring of resources such as CPU, memory, storage, network, state of essential services etc. Next level is applications and chained applications. Examples, monitoring mail queues of a mail server or databases from a SQL Server and so on.

Today, many monitoring products are available; many of these are tailored to special purposes and don’t cover all aspects of an Infrastructure.
I have noticed in recent years that many organizations are searching for a single product that can be used for monitoring as many components. Because nowadays most organizations run workloads on virtualized infrastructure, this means an extra challenge for the monitoring software.

Years ago, when I worked as a Sysadmin and virtualization was in a very early stage, my favorite monitoring software was a combination of the following products Nagios, Cacti and an advanced Syslog server.
Nagios has its origins as an Open Source product. Due to its open source nature, there are many products derived from Nagios, examples; OP5, Opsview, Groundwork, Check_MK and many more.

Some time ago Check_MK caught my attention, mainly because of its versatility and its ability to monitor diverse infrastructure including VMware vSphere.

In this and subsequent blog posts, I will investigate the potential of Check_MK, in particular the possibilities to monitor vSphere and other VMware products. Since there are already many excellent articles written about the installation and configuration I will not repeat these steps. Where needed, I will include references to articles that I used to build my Proof-of-Concept and issues that I encountered.

Read the rest of this entry »