Delve for android

… is useless – at least for me. We’re not an „all Microsoft“ kind of organisation. This way, it just shows if somebody had the time to save documents on our corporate sharepoint site. Yay…

[X] Uninstall

Backup!

Backups are a good thing to have (although I keep screwing up the server I’m doing the backups on, but that’s another story).

For my Ubuntu systems I found „backintime“ which does a periodic backup and sends it to a local directory or server.

To keep things backed up on my NAS (Ubuntu based) I installed snapper, which does a snapshot of btrfs and zfs file systems. Be sure to add the right „ALLOW_GROUPS“ if you’re using it with Windows – that way, you can use the Windows „Previous Versions“ GUI to access previous versions on your samba network share.

„Dependency hell“ all over

Ubuntu MAAS, Snaps, LXD containers, conjure-up – just a few new names and packages within the „Ubuntu Universe“ to create, manage and handle cloud infrastructure.

It’s. A. Mess. Everything seems to be designed to run on it’s own, empty, fresh install of Ubuntu-something, automatically creating, installing, destroying things. And instead of a library/ package dependency hell, you are rewarded with a mount and network service dependency hell. And often with a „something failed, look at $logfile“ whereas $logfile tells you, that super-service-whatthefuckd could not be started since the address/ port is already in use.

It works!

20 (16)TB of hard disk space:

Balance on '/data' is running
289 out of about 432 chunks balanced (290 considered), 33% left
Label: none uuid: 5fb8255b-5201-44af-8712-6c8dff56f033
 Total devices 5 FS bytes used 666.82GiB
 devid 1 size 3.64TiB used 425.01GiB path /dev/sde1
 devid 2 size 3.64TiB used 425.01GiB path /dev/sdc1
 devid 3 size 3.64TiB used 288.00GiB path /dev/sdg1
 devid 4 size 3.64TiB used 288.00GiB path /dev/sdd1
 devid 5 size 3.64TiB used 288.00GiB path /dev/sdf1

Data, RAID0: total=270.00GiB, used=258.90GiB
Data, RAID5: total=1.12TiB, used=406.78GiB
System, RAID1: total=8.00MiB, used=128.00KiB
Metadata, RAID1: total=2.00GiB, used=1.14GiB
GlobalReserve, single: total=512.00MiB, used=0.00B

Coming up next: automatically scaling wordpress cluster (not that I’m going to ever need it)…

systemd enables your swap (even if you don’t want it to)

Traditionally, swap space is configured in /etc/fstab. It can be turned off with „swapoff -a“ (removes all swap devices).

Enter systemd (on Ubuntu 16.04): „swapoff -a“ turns of the swap – but only for 30 seconds. If you want to deactivate your swap „more permanently“, uncomment the swap part of /etc/fstab. You think you’re done? Think again, because „systemd knows best“ and goes ahead and reactivates all swap partitions (as it seems, that are on the same disc as /, didn’t test that yet) regardless your settings in fstab.

Given enough time, systemd rants are a good place to learn about non-expected systemd behaviour.

OData – dead?

I am working with a software (TM1) that shifts it’s traditional, C-Based API to a REST-based, OData „compatible“ API. The API description is based on an XML file, the API calls are accepting/ emmitting JSON files.

The thing is: All exploration tools and services seem to be discontinued. The XOData Explorer is based on a chrome web app that is discontinued. The reference OData Explorer only understands the API version V3 (V4 is the latest standard). The metadata of the service I’m trying to use can not be viewed with the Visual Studio 2017 OData Connected Services. The proxy class generator only works with Visual Studio 2015 and older. The examples were hosted on CodePlex, which Microsoft discontinued in favor of github.

It doesn’t seem to be smart to develop a toolbox for the API based on the smell of rot concerning OData. But then, what choice do I have?