Tiny Deathstars of Foulness

Hey look, there’s a new category on the Jamf Marketplace, available at,selecting the AppConfig category. The new AppConfig category gives administrators of any MDM that supports AppConfig access to a set of apps that support AppConfig. If you have an app that isn’t listed here, feel free to let me know. 

What does this mean? Well, AppConfig is a way of sending data into an app. App config allows a customer to deploy settings into applications on iOS devices in much the same way that settings can be sent into a Mac app via the defaults command. This means an end user could get an app installed on their device from the iOS App Store, a custom app, or a B2B app and that app would have any settings the user might need to connect to servers or configure the experience.

So what is Managed App Config? At it’s most basic, you identify a label and a value in XML and send it to an iOS device that’s running iOS 7 or later (e.g. via Jamf 9 and up). The vendor who makes the app has to basically define what those settings are. Which brings up an interesting problem never fully addressed with defaults domains: standardization and ease-of-use (although MCX was close).  is a consortium of MDM vendors and software vendors that maintain the emerging AppConfig standards around Managed App Config (within the confines of what Apple gives vendors) and then makes a feed of settings for apps that conform to those standards. Jamf is a founding member of, along with MobileIron and AirWatch. Examples of what you could put into the feed include 
  • Enabling certain features of apps
  • Server URLs
  • Logos (if they’re pulled dynamically)
  • Text labels
  • Language packs

To see a list of apps that are available, check out 

Managed App Config options are set by vendors at compile time within the code and then the XML sent with the app is parsed by the app at installation time. If you’re a software vendor who wants to get started with AppConfig, check out the Spec Creator from Jamf Research or get in touch with the developer relations team from any MDM vendor.

If you’re a customer of an app and would like to leverage Managed App Config and your vendor isn’t listed on the site, get in touch with them, as this is the future of app management and chances are that you won’t be the only organization looking to unlock this type of feature. 

Let’s look at how this actually works. The Managed App Config options per supported app are available on a feed. The feed is available at Here, as follows, you’ll see a list of all of the apps supported.

You can then copy the path for an app, such as com.adobe.Adobe-Reaser/1/appconfig.xml and append it to the end of the URL to get the feed for that specific app. You can test this using to see output as follows.

Here, note that most of these fields are key value pairs defined by Adobe (in this example at least). You can enable or disable features of Adobe Reader using these keys. The same is true with a tool like Box that might want a more granular collection of settings than a feature like Managed Open In. 

Once you have the XML, you can then copy it to the clipboard and paste it into the App Configuration tab of an app, as follows. 

Finally, Apple has sample code available at

March 13th, 2018

Posted In: iPhone, JAMF

Tags: , , , , , , ,


The past couple of years has forced me to rethink many of my recommendations for how you backup computers in small office and home environments. Previously, I would have said that you could use a disk attached to an Apple AirPort. But the AirPort Base Station is no longer being made. Previously, I would have said you could use Time Machine Server, a service built into macOS Server in 5.4 and below. But that service is no longer being made in macOS Server by Apple and is now found in the Sharing System Preference pane . Previously, I might have even said to use the home edition of CrashPlan, which could have backed up to their cloud and/or a home server. But that plan is no longer being offered by Code 42.

So what are we to do? Well, luckily now the offerings out there are just endless. One of those offerings is so easy, you can run out to Best Buy, return home with a WD (Western Digital) drive, and be up and running in about 5 minutes. I’ll cover other options when I cover file services and Synology. But in the meantime, let’s look at setting up a WD drive, account, and configuring both to work with Time Machine. 

Setup Your WD Hard Drive
First, we’ll setup the drive. This is pretty straight forward. Plug the ethernet cable into your network, wait for the drive to boot up, and then go to the MyHome setup page.

Here, you’ll be prompted to setup a My Cloud Home account. Enter a name, email address, and password. Then click on Create Account.

You’ll then be prompted for the device you plugged in, which is discovered on the network. Click Connect.

Choose whether you want to share product improvement data. Ever since my team as a product manager I’m a huge fan of doing so, so I clicked Share.

Once that’s done, you’ll be prompted to get the desktop app. While not absolutely necessary, it’s not a bad idea. If you want the app, click Download.

Once the app is done downloading, open the directory and open the installer.

Click Install Now.

Once complete, you’ll see the menu bar. Click it and then add your device if you don’t see it by clicking on “I don’t see my device” 

When prompted, enter your email address and password that you created earlier and then click on Sign In.

Click Skip.

Next, in the notifications area for updating the software make sure to run that. There was a pretty bad vulnerability awhile back and that will make sure you’re good. Then click on the name of your WD MyCloud Home.

Add IFTTT Alerts

I want to see when new updates, channels or options are added, so I’m going to enable that. To do so, click on Services in the sidebar. and then click on Enable for IFTTT.

Assuming the terms of service are acceptable, click “I Agree”

When prompted, choose to connect to IFTTT.

From the IFTTT site, click Connect.

Choose which options to give IFTTT for the MyCloud API.

Browse the channels and enable each that you’d like and then click “Turn on.”

Mount the MyCloud Drive
Next, open a “Connect to Server” dialog box (Command-K from the Finder) and click on Browse.

Click on the MyCloud-XXX where XXX is the identifier for your MyCloud account.

Click on the timemachinebackup folder.

The folder should initially be empty. Now let’s open the Time Machine System Preference pane.

Click on “Select Backup Disk…”

Choose Your MyDisk From Time Machine

Choose the TimeMachineBackup directory for the MyCloud Device and click on “Use Disk.”

You’ll then want to create a user for backing up. To do so, go back to the site and click on settings. Then click on “Add user…” and enter an email address.

The email address will get an email to setup an account. Do so and then once you’ve configured the user, enter the email address and password when prompted.

Now wait for the first backup to finish. If you ever see any errors, check them; otherwise, you should backup to the device as with a locally attached drive, but you won’t need to plug directly into the drive to run backups.

This doesn’t solve for a lot of use cases that Time Machine Server would have been better for. But it’s a simple task that should cost you a little over a hundred bucks and get you backing up. I’m still a fan of cloud services. Backblaze, Carbonite, and others will backup your data for an annual fee of a little less than what a MyDrive costs. I’ll cover those in later articles, but for now, you’ve got a backup on your network, which even if you use one of those services is a great option in the event of hardware failure, as you can quickly get back up and running with a full system restore!

March 12th, 2018

Posted In: Mac OS X, Network Infrastructure

Tags: , , , , , ,


I have a new article for Thrive Global (another Arianna Huffington property) available at Thrive Global. This one is on “Tools and best practices on monitoring and teaching your kids responsible mobile device use.” It starts out like this:
My world changed when I awoke one day to find my 4-year-old daughter with a tablet in her hands, watching Transformers. The sight unleashed a handful of worries I hadn’t before experienced. Prior to that morning, I knew her to be fan of Star Wars figures, Legos and stuffed animals. And while I wasn’t displeased by her choice to watch a Michael Bay movie, I did start thinking about what else she could access on the device.
Click here to read more…
Screenshot of "Embracing (and managing) tech for your iGen child"

March 11th, 2018

Posted In: iPhone

Tags: , , , , ,

Leave a Comment

In an earlier article, I mentioned that MAMP Pro was still the best native GUI for managing web services on the Mac, now that macOS Server will no longer serve up those patchy services. After we cover the management in this article, you’ll likely understand why it comes it at $59. 

So you’ve installed MAMP. And you need more than the few basic buttons available there. So MAMP Pro came with it and you can try it for a couple of weeks for free. When you open MAMP Pro, you’ll see a screen where you can perform a number of management tasks. This is a more traditional side-bar-driven screen that will look like what Server Admin might have looked like before the web services screen got simplified in macOS Server.

The Hosts item in SETTINGS will show you each host installed on the server. Think of a host as a site. Each web server can serve up a virtually unlimited number of websites. You can configure an IP binding to the site, or hav
If you click on the plus sign, you can add a site. In this example, I’ll add and then click on create. When doing so, you can configure a database for each site (e.g. if you’re doing multi-tenant hosting), build a site off a template, or select a root directory for the site. 

The Apache tab of each host allows you to configure host-specific settings, including enabling options for directives such as Indexes, Includes, SymLink following, and CGI. More options than were in macOS Server for sure. You can also order allows, allow overrides, add new directives, set the index (or the default page of each site), add additional virtualhosts (such as for, and add a server admin email address. 

These were Apache-centric settings for each host. Click on the Nginx tab if you’re using Nginx instead of Apache. Nginx is a bit less “patchy” so there are a fewer options here. But they’re similar: Configure an index, add parameters, and a feature not available in the GUI options for Apache: allow or deny access based on IP.
The SSL tab allows you to generate a CSR, upload the cert and key file, and force connections to use https.

The Extras tab allows you to automatically install standard web packages. For example, here we’ll select WordPress.

Click on the Databases tab. To connect a site to a database, enter the name of the database when prompted. Note: the site itself will need credentials in order to connect, and if you’ve setup an “Extra” in the above step, the database will automatically be configured.

Next, let’s configure the ports used by the web servers. The previous settings were per-site. The rest that we cover in this article will be per-server, as these are global settings applied to the daemons themselves. Each of those services will have a port or ports associated with them. For example, the standard web port used is 80 or 443 for SSL-based connections and the standard port for MySQL is 3306. For publicly-facing sites these would be the standard ports, and given how common they are, there’s a button for “Set ports to 80, 81, 443, 7443, 3306”. Otherwise, you can enter each independently. Because the attaching of daemons is done here, this is also where you configure the user that services run as, as well as when to start the services and truncate log files.

The Editor option configures how the editor appears, which we’ll cover last in this article. The Editing option manages how the editor works (e.g. things  like tabs, autocompletes, etc.

The Fonts & Colors tab allows you to select each color assigned to various types of text.  

The Default Apps tab allows you to configure which app is opened when opening each type of file supported. 

Again, we’ll look at the editor later in this article. First, let’s finish getting the web server setup. Click on Apache. Here, you can load new Apache mods you download from the interwebs. I should mention that an important security step in locking down a publicly-facing web server is to disable all of the mods you don’t absolutely need. 

At the bottom of this screen, there’s also a handle little link to the directory with your logs, so you can read through them if needed.

The Nginx option underneath is similar. Access to log files is there, as is the ability to enable installed Nginx mods. 

The MySQL option also provides access to some straight-forward command-line options, but in a nice GUI. Here, you can configure a root password for MySQL ( which does this: Reset A Lost MySQL Password ), enable phpMyAdmin, MySQL Workbench, and Sequel Pro-based administration, enable network access to the MySQL Service (using ports configured in the Ports section of the app) which I cover at Allow Remote Connections To MySQL, and view logs.

The Dynamic DNS options are cool. Click there, and if your web server is behind a DHCP address, you can configure a dynamic DNS service including DNS-O-Matic,,,, etc. This way when you reboot and get a new IP address from your ISP, it’ll update the service automatically.

Memcached is a distributed memory object caching system. It’s used to make sites appear faster or to distribute caching between servers for systems that, for example, get clustered. It’s included here for a reason, I’m sure of it! Either way, I actually use it for a few things and like the fact that it’s there. To enable, simply choose how much memory to give it, configure the logging level (usually low unless you’re troubleshooting), and gain access to logs. If you check the “Include Memcached server in GroupStart” then memcache will fire up when you start your web services.

Click postfix. Here, you configure your server to route mail through an email account. If you run this from the command line, you can also configure your server to be a mail server; however, when you do that you’re likely to get mail bouncing all over the place. So if the server or a service on the server is supposed to send mail, it’s usually best to route through something like a gmail account. 

The Languages section allows you to configure how PHP, Python, Perl, and Ruby work on the server. For PHP, you can configure which version of PHP is installed, configure a version of PHP for hosts, enable caching (different than memcached), enable a few basic extensions (I’ve been playing with oauth a lot recently), choose logging options, and have a simple way to see the logs. 

Since you’re running on a Mac, you already have Python, but if you click on the Python option, you can make the version of Python bundled with Mac is 2.7.10 instead of 2.7.13.

Click on Perl to do the same.

Click on Ruby to do the same.

The editor is also pretty easy to use. Simply use the plus sign to add a file you’d like to edit. Keep in mind when browsing that everything MAMP Pro needs is self-contained in the /Applications/MAMP directory, so it should be pretty easy to find files for editing. 

And that’s it. This seems like a lot of stuff, but between sites like ServerFault and other Apache/Nginx articles, you’ll likely find most of the things you need. It’s worth mentioning that I consider this another baby step to just managing Apache using config files. macOS Server tried hard to reduce the complexity of where different settings and options are derived from; MAMP Pro makes no allusion that web server management should be so simple. That’s one of the things I like about it. It’s like you went from riding in a buggy on the back of a bike to riding with training wheels. The more you know, the better off you are.

March 10th, 2018

Posted In: Mac OS X, Mac OS X Server, Mac Security, WordPress

Tags: , , , , , , , , , , ,


February 21st, 2018

Posted In: MacAdmins Podcast

Tags: , , ,

Been working on a new plugin to embed device details from Jamf Pro into Jira Service Desk. It looks a little like this:

To access the plugin, see the links below.

February 13th, 2018

Posted In: JAMF

Tags: , , , , ,

Many of the people that read my articles undoubtedly arleady know this, but Apple has announced a sharp reduction in the number of services provided. Per this article, the Calendar, Contacts, DHCP, DNS, Mail, Messages, NetInstall, VPN, Websites, and Wiki services are being deprecated and Apple has provided a few services, per service, that they recommend moving to. Those services, per the above article, include the following:











I’ve been saying many of these services/features should go away in macOS Server so the developers could focus on providing an excellent experience and solid QA/unit testing for the services/features that remain. The fact that apps are being swiftified is great, as it speaks volumes to the future of the services themselves. The fact that Apple is reducing the number of licenses they’re tracking and the mistake they’re allowing customers to make is also great.

Having said that, every time I think that a service should go away, I hear from someone that they rely on that service. Most of this feedback comes from consultants who have made the server a central part of their consultancy. As someone who used to plan services as products for customers in consultancies, if you find yourself in similar situations when planning where services go when Apple retires them, I would strongly recommend looking at SaaS solutions where customers can give you a login and you can help guide them into a new and better solution. At least, that’s the way I positioned most of these services in the last version of the macOS Server book…

Yes, it was great having Apple handle all of the patching and customers were able to take advantage of a lot of technology with very few resources. However, that’s just not where we are any more. And rather than argue about it or try emailing Tim Cook or make petitions or even complain, save your cycles and look for new and better replacements for each service (preferably not ones that require physical servers, provided that customers are okay with that)! 

And stay tuned. I suspect we’ll cover this on an upcoming episode of the Mac Admins Podcast! 😉

What are your thoughts? Remorse? Applause?

January 25th, 2018

Posted In: Mac OS X, Mac OS X Server

Tags: , , ,

December 29th, 2017

Posted In: MacAdmins Podcast

Tags: , , , , ,

My latest Huffington Post article, called 20 Constants In Software Development is up. It starts out like this:

There are so many things I wish people had told me when I was in school, or earlier in my career. Things that aren’t variable between organizations you work with, or even teams you work in. So I thought I’d jot a few down of these for software development teams (if only to prove that no, despite what product managers say, you aren’t crazy). So here goes:

  1. A project will never have enough people to build all the features you want. Period.
  2. Less features means fewer defects.
  3. As a software project nears completion the amount of work remaining rises in proportion to how many hacks and shortcuts you took.
Read more…

December 11th, 2017

Posted In: Articles and Books, Programming

Tags: , , ,

High Sierra sees the Caching service moved out of macOS Server and into the client macOS. This means administrators no longer need to run the Server app on caching servers. Given the fact that the Caching service only stores volatile data easily recreated by caching updates again, there’s no need to back the service up, and it doesn’t interact with users or groups, so it’s easily divested from the rest of the Server services.

And the setup of the Caching service has never been easier. To do so, first open System Preferences and click on the Sharing System Preferences pane.

From here, click on the checkbox for Content Caching to start the service.

At the Content Caching panel, the service will say “Content Caching: On” once it’s running. Here, you can disable the “Cache iCloud content” option, which will disable the caching of user data supplied for iCloud (everything in here is encrypted, by the way). You can also choose to share the Internet Connection, which will create a wireless network that iOS devices can join to pull content. 

Click Options. Here, you can see how much storage is being used and limit the amount used. 

defaults read /Library/Preferences/

Which returns the following configurable options:

Activated = 1;
CacheLimit = 0; DataPath = “/Library/Application Support/Apple/AssetCache/Data”; LastConfigData = <BIGLONGCRAZYSTRING>; LastConfigURL = “”; LastPort = 56452; LastRegOrFlush = “2017-09-11 16:32:56 +0000”; LocalSubnetsOnly = 1; PeerLocalSubnetsOnly = 1; Port = 0; Region = 263755EFEF1C5DA178E82754D20D47B6; ReservedVolumeSpace = 2000000000; SavedCacheDetails = {
SavedCacheSize = 0;
ServerGUID = “EB531594-B51E-4F6A-80B9-35081B924629”;
Version = 1;}

This means that all those settings that you used to see in the GUI are still there, you just access them via the command line, by sending defaults commands. For example, 

defaults write /Library/Preferences/ CacheLimit -int 20000000000

You can

AssetCacheManagerUtil status

Which returns something similar to the following:

2017-09-11 11:49:37.427 AssetCacheManagerUtil[23957:564981] Built-in caching server status: {
Activated = 1;
Active = 1;
CacheDetails = {
iCloud = 4958643;
“iOS Software” = 936182434;};
CacheFree = 472585174016;
CacheLimit = 0;
CacheStatus = OK;
CacheUsed = 941141077;
Parents = ();
Peers = ();
PersonalCacheFree = 472585174016;
PersonalCacheLimit = 0;
PersonalCacheUsed = 4958643;
Port = 56452;
PrivateAddresses = (“”);
PublicAddress = “”;
RegistrationStatus = 1;
RestrictedMedia = 0;
ServerGUID = “EB531594-B51E-4F6A-80B9-35081B924629”;
StartupStatus = OK;
TotalBytesDropped = 0;
TotalBytesImported = 4958643;
TotalBytesReturnedToChildren = 0;
TotalBytesReturnedToClients = 166627405;
TotalBytesReturnedToPeers = 0;
TotalBytesStoredFromOrigin = 166627405;
TotalBytesStoredFromParents = 0;
TotalBytesStoredFromPeers = 0;

You can also use AssetCacheManagerUtil to manage tasks previously built into the Server app. To see the available options, simply run the command:

bash-3.2# /usr/bin/AssetCacheManagerUtil

Which would show the following:

Options are:
-a|–all show all events
-j|–json print results in JSON
-l|–linger don’t exit
2017-09-11 11:57:30.066 AssetCacheManagerUtil[24213:569932] Commands are:
moveCacheTo path
absorbCacheFrom path read-only|and-destroy

As such, to enable the server:

bash-3.2# /usr/bin/AssetCacheManagerUtil activate 

To disable the server

bash-3.2# /usr/bin/AssetCacheManagerUtil deactivate

To check if the server can be activated

bash-3.2# /usr/bin/AssetCacheManagerUtil canActivate

To flush the cache of assets on the server:

bash-3.2# /usr/bin/AssetCacheManagerUtil flushCache 

To reload settings if you make any changes:

bash-3.2# /usr/bin/AssetCacheManagerUtil reloadSettings

To move the database

/usr/bin/AssetCacheManagerUtil moveCacheTo "/Volumes/SONY/Library/Application Support/Apple/AssetCache/Data"

Finally, if you’d like to see the caching server your client system is using, you can run the following command:

/usr/bin/AssetCacheLocatorUtil 2>&1 | grep guid | awk '{print$4}' | sed 's/^\(.*\):.*$/\1/' | uniq

And if you use Jamf Pro and would like to use this as an extension attribute, that’s posted here: I didn’t do any of the if/then there, as I’d usually just do that on the JSS.

Note: To see how AssetCache interacts with Tetherator, see Tethered Caching of iOS Assets from macOS 10.12.4.

September 28th, 2017

Posted In: Mac OS X, Mac OS X Server

Tags: , , , , , ,

Next Page »