DDEV is continually focused on quick project startup and fast responses to its web requests. DDEV’s performance is mostly an issue of how Docker runs on your workstation.
On Linux, including Windows WSL2 and Gitpod, Docker is fast. Most people are happy with Linux performance and don’t need to change anything.
On macOS and Windows with Docker Desktop, allocated resources and mounted filesystem performance can be significant bottlenecks. Taking a bit of time to optimize your setup can yield massive performance gains.
Mutagen can offer a big performance boost on macOS and Windows. It’s fast and doesn’t need any setup; you only need to enable it. Before Mutagen, Mac and Windows users configured NFS for speed improvements—though it requires setup and isn’t as fast.
If you’re on macOS or Windows, you can skip everything else here and enable Mutagen:
What Mutagen Does¶
The Mutagen asynchronous caching feature is the best way to improve DDEV’s web-serving performance on macOS and Windows, and we recommend it for most projects. It can be significantly faster than NFS, massively faster than plain Docker or Colima, and it makes filesystem watchers (
inotify) work correctly.
Mutagen decouples in-container reads and writes from reads and writes on the host machine, so each can enjoy near-native speed. A change on the host gets changed “pretty soon” in the container, and a change in the container gets updated “pretty soon” on the host; neither filesystem is stuck waiting on the other one. This “pretty soon” means, however, that there’s a brief window where files on the host may not exactly match the files inside the container—so files that manage to change in both places can lead to conflicts.
Docker bind-mounts, the traditional approach to getting your code into DDEV’s web container, check every file access against the file on the host. Docker’s way of doing these checks macOS and Windows is not very performant, even with NFS. Linux and Linux-like systems are faster because Docker provides native file-access performance.
While Mutagen works fine and has automated tests for Linux and Windows WSL2, it may not be worth enabling on those systems since it won’t make the dramatic difference it does on macOS and Windows.
Another major advantage of Mutagen over NFS is that it supports filesystem notifications, so file-watchers on both the host and inside the container will be notified when changes occur. This is a great advantage for many development tools, which otherwise have to poll for changes at greater expense. Instead, they can be notified via normal
Enabling and Disabling Mutagen¶
Don’t Install Mutagen
Don’t install the
mutagen binary separately. If it’s not available, DDEV will install and upgrade it for you.
We recommend enabling Mutagen globally with
ddev config global --mutagen-enabled. You can disable it again with
ddev mutagen reset && ddev config global --mutagen-enabled=false.
If you’d rather use Mutagen on a specific project, run
ddev stop, enable it with
ddev config --mutagen-enabled, and
ddev start again.
You can’t disable Mutagen on individual projects if it’s enabled globally.
To stop using Mutagen on a project, run
ddev mutagen reset && ddev config --mutagen-enabled=false.
nfs-mount-enabled feature is automatically turned off if you’re using Mutagen.
Mutagen and User-Generated Uploads¶
When Mutagen is enabled, DDEV attempts to exclude user-generated files in
upload_dir—when it exists—from syncing. It does this by using a bind-mount in the generated docker-compose configuration, and excluding the directory from syncing in
.ddev/mutagen/mutagen.yml. This behavior is automatic and you shouldn’t have to take any action in most cases.
If you have a non-standard location for user-generated files, like
private/fileadmin with the deprecated
typo3-secure-web approach, you should override the project defaults by setting
.ddev/config.yaml and pointing it at the correct directory. This will allow Mutagen to sync correctly.
Mutagen Integration Caveats¶
If you run into an issue with Mutagen, please try to recreate and report it via a support channel!
Mutagen has generally been great for those using it, but it’s good to be aware of its trade-offs:
- It’s not the right choice for every project.
Filesystem consistency has been excellent with Mutagen, but performance is its specialty. If consistency is your highest priority, then there are reasons to be cautious. Two-way sync is a very difficult computational problem, and problems may surface.
- Avoid file changes when DDEV is stopped.
If you change files—checking out a different branch, removing a file—while DDEV is stopped, Mutagen has no way to know about it. When you start again, it will get the files that are stored and bring them back to the host. If you do change files while DDEV is stopped, run
ddev mutagen resetbefore restarting the project so Mutagen only starts with awareness of the host’s file contents.
- It modestly increases disk usage.
Mutagen integration increases the size of your project code’s disk usage, because the code exists both on your computer and inside a Docker volume. (As of DDEV v1.19+, this does not include your file upload directory, so normally it’s not too intrusive.) Take care that you have enough overall disk space, and that on macOS you’ve allocated enough file space in Docker Desktop. For projects before DDEV v1.19, if you have a large amount of data like user-generated content that doesn’t need syncing (i.e.
fileadminfor TYPO3 or
sites/default/filesfor Drupal), you can exclude specific directories from getting synced and use a regular Docker mount for them instead. As of v1.19, this is handled automatically and these files are not Mutagen-synced.
- Beware simultaneous changes to the same file in both filesystems.
As we pointed out above, any project likely to change the same file on the host and inside the container may encounter conflicts.
- Massive changes can cause problems.
Massive file changes on the host or in the container are the most likely to introduce issues. This integration has been tested extensively with major changes introduced by
ddev composer create, but be aware of this issue. Changing Git branches,
yarn install, or a script that deletes huge sections of the synced data are related behaviors that should raise caution. Again, use
ddev mutagen resetbefore restarting the project if you want to be sure Mutagen starts out looking at the host machine’s files.
- Mutagen is asynchronous.
A massive change in either filesystem can result in lag as all changed files are handled. You can use
ddev mutagen monitorto get a better look at what’s happening.
- You can manually trigger a sync.
ddev stopautomatically force a Mutagen sync. You can cause an explicit sync with
ddev mutagen syncand see syncing status with
ddev mutagen status.
- Be mindful of in-container Composer actions.
If you run Composer actions inside the container with
ddev ssh, it’s a good idea to run
ddev mutagen syncand make sure they’re synced as soon as possible. Most people won’t notice the difference and Mutagen will get it synced soon enough.
- Perform big Git operations on the host side.
Git actions that change lots of files, like switching branches, are best done on the host side and not inside the container. You may want to do an explicit
ddev mutagen synccommand after doing something like that to be sure all changes are picked up quickly.
- Share projects carefully with non-Mutagen users.
If you share a project with some users that want Mutagen, perhaps on macOS, and other users that don’t want or need it, perhaps on WSL2, don’t check in
mutagen_enabled: true. Instead, either use global Mutagen configuration or add a not-checked-in, project-level
.ddev/config.mutagen.yamljust to include
mutagen_enabled: truein it. That way, only users with that file will have Mutagen enabled.
- Windows symlinks have some Mutagen restrictions.
On macOS and Linux (including WSL2) the default
posix-rawtype of symlink handling. (See mutagen docs). This basically means any symlink created will try to sync, regardless of whether it’s valid in the other environment. Mutagen, however, does not support
posix-rawon traditional Windows, so DDEV uses the
portablesymlink mode. The result is that on Windows, using Mutagen, symlinks must be strictly limited to relative links that are inside the Mutagen section of the project.
- It’s a filesystem feature. Make backups!
If we’ve learned anything from computer file-storage adventures, it’s that backups are always a good idea!
In general, it’s best practice on most projects to do significant Git operations on the host, but they can be disruptive to the sync. It’s easy to add a Git post-checkout hook to do a
ddev mutagen sync operation though. Add a
.git/hooks/post-checkout file to your project and make it executable with
chmod +x .git/hooks/post-checkout:
Actions by those programs can also set off massive filesystem changes.
You should run
ddev mutagen sync in order to get things into sync, or simply wait.
Advanced Mutagen Configuration Options¶
The Mutagen project provides extensive configuration options.
Each DDEV project includes a
.ddev/mutagen/mutagen.yml file by default, with basic defaults you can override if you remove the
#ddev-generated line at the beginning of the file.
If you edit the
- Remove the
- Execute a
ddev mutagen resetto avoid the situation where the Docker volume and Mutagen session still have files from an older configuration.
The most likely thing you'll want to do is to exclude a path from mutagen syncing, which you can do in the
paths: section of the
ignore: stanza in the
You can exclude a path from Mutagen syncing and bind-mount something from the host or a different volume on that path with a
docker-compose.*.yaml file. So if you have a heavy project subdirectory (lots of fonts or user-generated content, for example), you could exclude that subdirectory in
.ddev/mutagen/mutagen.yml and add a
For example, if you want the
stored-binaries subdirectory of the project to be available inside the container, but don’t need Mutagen to be syncing it, you can use normal Docker bind-mounting for that subdirectory:
- Take over the
.ddev/mutagen/mutagen.ymlby removing the
/stored-binariesto the excluded paths:
- Add a
Troubleshooting Mutagen Sync Issues¶
- Please make sure that DDEV projects work without mutagen before troubleshooting mutagen.
ddev config --mutagen-enabled=false && ddev restart.
export DDEV_DEBUG=truewill provide more information about what's going on with mutagen.
- As of DDEV v1.21.2, DDEV's mutagen daemon keeps its data in a DDEV-only MUTAGEN_DATA_DIRECTORY,
DDEV's private mutagen binary is installed in
~/.ddev/bin/mutagen. You can use all the features of mutagen with
export MUTAGEN_DATA_DIRECTORY=~/.ddev_mutagen_data_directoryand running the mutagen binary in
~/.ddev/bin/mutagen, for example:and
~/.ddev/bin/mutagen daemon stop.
You can run the diagnose_mutagen.sh script to gather information about Mutagen’s setup. Please share output from it when creating an issue or seeking support.
~/.ddev/bin/mutagen daemon stop && ~/.ddev/bin/mutagen daemon startto restart the Mutagen daemon if you suspect it’s hanging.
ddev mutagen resetif you suspect trouble, and always after changing
.ddev/mutagen/mutagen.yml. This restarts the project’s Mutagen data (Docker volume + Mutagen session) from scratch.
ddev mutagen monitorcan help watch Mutagen behavior. It’s the same as
~/.ddev/bin/mutagen sync monitor <syncname>.
ddev debug mutagenwill let you run any Mutagen command using the binary in
- If you’re working on the host and expecting things to show up immediately inside the container, you can learn a lot by running
ddev mutagen monitorin a separate window as you work. You’ll see when Mutagen responds to your changes and get an idea about how much delay there is.
ddev stopbefore massive file change operations, like moving a directory.
- If you get in real trouble, run
ddev stop, reset your files with Git, and run
ddev mutagen resetto throw away the Docker volume which may already have incorrect files on it.
Advanced Mutagen Troubleshooting¶
Most people get all the information they need about mutagen by running
ddev mutagen monitor to see the results. However, Mutagen has full logging. You can run it with
ddev mutagen logs.
Mutagen Strategies and Design Considerations¶
Mutagen provides enormous speed boosts in everyday usage, but it’s trying desperately under the hood to keep container changes reflected on the host and vice versa.
DDEV mounts a fast Docker volume onto
/var/www/html inside the
web container, then delegates to the Mutagen daemon (on the host). That daemon has the job of keeping all host project contents in sync with the contents of the Docker volume.
Consistency is a high priority for DDEV’s integration, which tries at key points to make sure everything is completely consistent and in sync.
The Mutagen daemon’s life cycle and sync sessions are something like this:
ddev start, the Mutagen agent will be started if it’s not already running.
- If there’s already a sync session for the project, it’s resumed.
ddev pause, the sync session is flushed to ensure consistency, then paused.
A synchronous flush happens after any
ddev composer command, since Composer may cause massive changes to the filesystem inside the container that need to be synced before operation continues.
If you need to reset everything for a project, you can do it with
ddev mutagen reset, which starts the Mutagen session from scratch and removes the Docker volume so it can be recreated from scratch.
Safe to Use with Other Mutagen Installations¶
DDEV requires and provides a specific version of Mutagen, which you can see with
ddev version. If another
mutagen instance or daemon is installed on your workstation it doesn't matter, because DDEV's version runs separately and uses a different data directory.
Using NFS to Mount the Project into the Web Container¶
NFS (Network File System) is a classic, mature Unix technique to mount a filesystem from one device to another. It provides significantly improved web server performance on macOS and Windows. It doesn’t really impact performance on Linux, so we don’t recommend it there.
DDEV supports this technique and requires pre-configuration on your host computer, facilitated by a setup script that asks for your
sudo password when it’s necessary:
- Make sure DDEV is already working and you can use it.
- Use the script below for your OS to configure the NFS server and exports files.
- Test that NFS is working correctly by using
ddev debug nfsmountin a project directory. The first line should report something like “Successfully accessed NFS mount of /path/to/project”.
- Enable NFS mounting globally with
ddev config global --nfs-mount-enabled.
You can also configure NFS mounting on a per-project basis with
ddev config --nfs-mount-enabledin the project directory, but this is unusual. If NFS mounting is turned on globally, it overrides any local project settings for NFS.
ddev startyour project and make sure it works normally. Use
ddev describeto verify that NFS mounting is being used. The NFS status is near the top of the output of
Skip step 2 if you’re already using NFS!
If you’re already using NFS with Vagrant on macOS, for example, and you already have a number of exports, the default home directory export here won’t work—you’ll have overlaps in your
/etc/exports. Or on Windows, you may want to use an NFS server other than Winnfsd like the Allegro NFS Server.
The recommendations and scripts below are for getting started if, like most people, you don’t already use NFS.
Download, inspect, make executable, and run macos_ddev_nfs_setup.sh:
This one-time setup stops running DDEV projects, adds your home directory to the
/etc/exports config file that
nfsd uses, and enables
nfsd to run on your computer.
This shares your home directory via NFS to any NFS client on your computer, so it’s critical to consider security issues. You can make the shares in
/etc/exports more limited, as long as they don’t overlap. NFS doesn’t allow overlapping exports.
If your DDEV projects are set up outside your home directory, you’ll need to add a line to
/etc/exports for that share as well:
sudo vi /etc/exports.
- Copy the line the script just created (
/System/Volumes/Data/Users/username -alldirs -mapall=<your_user_id>:20 localhost).
- Edit to add the additional path, e.g:
/Volumes/SomeExternalDrive -alldirs -mapall=<your_uid>:20 localhost.
macOS Full Disk Access for Special Directories¶
If your projects are in a subdirectory of the
~/Desktopdirectories, or on an external drive, you must grant “Full Disk Access” privilege to
/sbin/nfsdin System Preferences → Security & Privacy → Privacy. In the Full Disk Access section, click the “+” and add
You should then see
nfsdin the list:
sudo nfsd restart.
- From a project directory, run
ddev debug nfsmountto confirm successful output.
macOS NFS Debugging¶
- Temporarily disable any firewall or VPN.
showmount -eto find out what’s exported via NFS. If you don’t see a parent of your project directory, NFS can’t work.
- If nothing is showing, use
nfsd checkexportsand look carefully for errors.
ps -ef | grep nfsdto make sure
sudo nfsd restart.
- Add the following to your
- Run Console.app and search for “nfsd” at the top. Run
sudo nfsd restartand read the messages carefully. Try running
ddev debug nfsmountin the problematic project directory:
$ ddev debug nfsmount Successfully accessed NFS mount of /Users/rfay/workspace/d8composer TARGET SOURCE FSTYPE OPTIONS /nfsmount :/System/Volumes/Data/Users/rfay/workspace/d8composer nfs rw,relatime,vers=3,rsize=65536,wsize=65536,namlen=255,hard, nolock,proto=tcp,timeo=600,retrans=2,sec=sys,mountaddr=192.168.65.2,mountvers=3,mountproto=tcp,local_lock=all,addr=192.168.65.2 /nfsmount/.ddev
The executable components required for Windows NFS,
nssm, are packaged with the DDEV Windows Installer in each release. If you’ve used the Windows installer, they’re ready to go.
winnfsd as a service, please download, inspect and run
windows_ddev_nfs_setup.sh created by the installer at
You can also download this directly from the GitHub repository) in a Git Bash session on Windows. If your DDEV projects are set up outside your home directory, you’ll need to edit
~/.ddev/nfs_exports.txt (created by the script) and restart the service with
sudo nssm restart nfsd.
On Windows 10/11 you’ll likely have to allow
winnfsd to bypass the Windows Defender Firewall. If you’re getting a timeout with no information after
ddev start, try going to Windows Defender Firewall → Allow an app or feature through Windows Defender Firewall, Change Settings, Allow another app. Then choose
C:\Program Files\ddev\winnfsd.exe, assuming that’s where
winnfsd is installed.
ddev start Failures with NFS Mount Enabled¶
There are a number of reasons the NFS mount can fail on
- Firewall issues.
- NFS Server not running.
- Trying to start more than one NFS server.
- Overlapping NFS exports, typically an issue if you’ve established another NFS client like Vagrant. You’ll need to reconfigure your exports paths so they don’t overlap.
- Path of project not shared in
To debug and solve permission problems:
ddev debug nfsmountin a project directory to see if basic NFS mounting is working. If that works, everything else probably will too.
- When debugging, run
ddev restartin between each change. Otherwise, you can have stale mounts inside the container and you’ll miss any benefit you may find in the debugging process.
- Restart the server with
sudo nssm restart nfsd.
Windows-specific NFS debugging¶
- Temporarily disable any firewall, VPN, or virus checker.
You can only have one NFS daemon running, so if another application has installed one, you’ll want to use that NFS daemon and reconfigure it to allow NFS mounts of your projects.
Stop the running
sudo nssm stop nfsd.
winnfsdmanually in the foreground with
winnfsd "C:\\". If it immediately returns to the shell prompt, there’s likely another
- In another window, in a DDEV project directory, run
ddev debug nfsmountto see if it can mount successfully. (The project doesn’t need to be started.) If
ddev debug nfsmountis successful, then everything is probably going to work.
~/.ddev/nfs_exports.txthas a line that includes your project directories, then run
sudo nssm start nfsdand
nssm status nfsd. The status command should show
- These nssm commands may be useful:
sudo nssm start nfsd,
sudo nssm stop nfsd,
nssm status nfsd,
sudo nssm edit nfsd(pops up a window that may be hidden), and
sudo nssm remove nfsd(also pops up a window, doesn’t work predictably if you haven’t already stopped the service).
nssmlogs failures and what it’s doing to the system event log. Run Event Viewer and filter events:
- Please make sure you’ve excluded
winnfsdfrom the Windows Defender Firewall per the installation instructions above.
- On Windows 10/11 Pro you can visit Turn Windows features on or off and enable Services for NFS → Client for NFS. The
showmount -ecommand will then show available exports on the current machine. This can help find out if a conflicting server is running or identifying a problem with exports.
Freeing Up System Resources¶
Every project you run uses system resources, and may compete for those resources. A reasonable practice is to individually stop projects you’re not using. You could also stop all projects with
ddev poweroff and only start the one you’re working on.
ddev list will display all your projects along with each one’s status.
Docker Desktop for Mac Settings¶
Open Docker Desktop’s Preferences, and visit Resources → Advanced. Here you can adjust the CPUs, memory, and disk allocated to Docker. The defaults work well for a small project or two, but you may want to adjust these upward based on your experience. Most people raise the memory allocation to 6GB or higher. The disk allocation almost always needs to be raised to accommodate increased downloaded images. Your experience will determine what to do with CPUs.