Albal's Blog

Sometimes a disk will drop out a RAID array, especially if you have drives that don’t support a low latency error recovery such as the Seagate Barracuda SFF drives.  You should notice an alarm in ESX or your operating system that says the Logical Drive has failed.  Associated with this will be a physical drive failure too. To recovery from the problem you will need to remove the affected drive for 10 seconds and then re-insert the drive into the slot it came from.  You can do all of this while the system is booted into the array management software. Once you have the physical drive recognised as operational again then you can make steps to recover the logical drive.  In my case I am using ESX but any other operating system that can use the hpssacli utility should be the same. First show the status of the logical drive:

ansible-playbook –check playbook.yml –diff –connection=local

Some simple powershell to show the used and provisioned size of disks in VMware

$report = @() foreach ($vm in Get-VM){ $view = Get-View $vm $row = ’’ | select Name, Provisioned, Actual, Thin $row.Name = $vm.Name $row.Provisioned = [math]::round($vm.ProvisionedSpaceGB , 2) $row.Actual = [math]::round(($vm.UsedSpaceGB ) , 2) $row.Thin = $view.config.hardware.Device.Backing.ThinProvisioned | Out-String $report += $row } $report | Sort Actual

vboxmanage list vms | cut -f 2 -d ‘"’ | xargs -I{} vboxheadless -s “{}”

mkdir -p ~/.vim/autoload ~/.vim/bundle && \
curl -LSso ~/.vim/autoload/pathogen.vim https://tpo.pe/pathogen.vim

git clone git://github.com/rodjek/vim-puppet.git ~/.vim/bundle/puppet

# ~/.vimrc
execute pathogen#infect()
syntax on
filetype plugin indent on
```

If you have automated systems sometimes it is useful to have them interact with your documentation from time to time.  Using the Confluence REST API you can upload an image as an attachment to the page and later update it if required.  If you follow the Atlassian documentation you end up getting a denied request for the update as it wishes for the new file to have a different filename. Below are the curl lines to post and then update the image.  Change your page ID and attachment ID as required:``` curl -D- -k -u $user:$pass -p -H “X-Atlassian-Token: nocheck” -X POST \ -F [email protected] https://confluence.domain.tld/confluence/rest/api/content/$page_id/child/attachment

Having trouble getting the latest Vagrant and VirtualBox working together because the ssh key is not acceptable to Vagrant?  Well it turns out that the ownership on authorized_keys is not 600 so it fails in operating systems like Centos.  The trick is to modify one of the files to add a chmod 0600 after the authorized_keys file has been updated.  Full information in the link: https://github.com/mitchellh/vagrant/issues/7610

I had a problem in that updates were not successful on the cluster.  Many reports for the error message “Upgrading a VMware ESXi host fails with the error: Cannot run upgrade script on host” point to removing the FDM installation using the uninstall.sh

/opt/vmware/uninstallers/VMware-fdm-uninstall.sh

However I didn’t have the uninstallers directory on my ESX boxes. My solution was to enable HA on the cluster then removed HA on the cluster and after that updates worked fine.

Ciphers aes128-ctr [email protected] aes192-ctr aes256-ctr [email protected] arcfour arcfour128 arcfour256 [email protected]  
MACs hmac-ripemd160 [email protected] [email protected] hmac-sha1 [email protected] hmac-sha2-256 [email protected] hmac-sha2-512 [email protected] [email protected] [email protected] [email protected] [email protected]

Buildroot is pure awesomeness for maintaining the build of a GNU/Linux based operating system.  However, during my ventures I’ve had to keep private code private so use a git repo behind ssh.  There is a way to achieve this with Buildroot but it is undocumented:

LIBFOO_SITE_METHOD = git
LIBFOO_SITE = ssh://[email protected]/group/repo.git
LIBFOO_VERSION = v1.01.a

On the first line we have to set the method to git as the auto detect won’t work on SSH
The second line contains the git repo as it would be defined if using git via ssh accept prepended with ssh:// and the the colon that is normally between the end of the FQDN and the location is replaced with a slash.
Line 3 can be a tag, branch or commit hash.