systemd中文文档

- 启动: http://systemd.cn/docs/Booting/
- 自动启动评估: http://systemd.cn/docs/Booting/AUTOMATIC_BOOT_ASSESSMENT/
- 启动组件和根文件系统探测: http://systemd.cn/docs/Booting/ROOTFS_DISCOVERY/
- 引导加载程序接口: http://systemd.cn/docs/Booting/BOOT_LOADER_INTERFACE/
- 恢复出厂设置: http://systemd.cn/docs/Booting/FACTORY_RESET/
- 挂载点可用性要求: http://systemd.cn/docs/Booting/MOUNT_REQUIREMENTS/
- TPM2 PCR测量: http://systemd.cn/docs/Booting/TPM2_PCR_MEASUREMENTS/
- Boot Loader Specification: http://systemd.cn/docs/BOOT_LOADER_SPECIFICATION/
- Discoverable Partitions: http://systemd.cn/docs/DISCOVERABLE_PARTITIONS/
- Elf Dlopen Metadata: http://systemd.cn/docs/ELF_DLOPEN_METADATA/
- Osc Context: http://systemd.cn/docs/OSC_CONTEXT/
- Package Metadata for Executable Files: http://systemd.cn/docs/PACKAGE_METADATA_FOR_EXECUTABLE_FILES/
- API File Systems: http://systemd.cn/docs/API_FILE_SYSTEMS/
- Appstream Bundle: http://systemd.cn/docs/APPSTREAM_BUNDLE/
- Backports: http://systemd.cn/docs/BACKPORTS/
- Booting Without /usr is Broken: http://systemd.cn/docs/SEPARATE_USR_IS_BROKEN/
- Code Quality Tools: http://systemd.cn/docs/CODE_QUALITY/
- Coding Style: http://systemd.cn/docs/CODING_STYLE/
- Compatibility with SysV: http://systemd.cn/docs/INCOMPATIBILITIES/
- Container Interface: http://systemd.cn/docs/CONTAINER_INTERFACE/
- Contributing: http://systemd.cn/docs/CONTRIBUTING/
- Control Group APIs and Delegation: http://systemd.cn/docs/CGROUP_DELEGATION/
- Converting Existing Users to systemd-homed: http://systemd.cn/docs/CONVERTING_TO_HOMED/
- Credentials: http://systemd.cn/docs/CREDENTIALS/
- Desktop Environment Integration: http://systemd.cn/docs/DESKTOP_ENVIRONMENTS/
- Diagnosing Boot Problems: http://systemd.cn/docs/DEBUGGING/
- File Descriptor Store: http://systemd.cn/docs/FILE_DESCRIPTOR_STORE/
- Frequently Asked Questions: http://systemd.cn/docs/FAQ/
- Governance: http://systemd.cn/docs/GOVERNANCE/
- Hacking on systemd: http://systemd.cn/docs/HACKING/
- Home Directories: http://systemd.cn/docs/HOME_DIRECTORY/
- Inhibitor Locks: http://systemd.cn/docs/INHIBITOR_LOCKS/
- Initrd Interface: http://systemd.cn/docs/INITRD_INTERFACE/
- Journal Export Formats: http://systemd.cn/docs/JOURNAL_EXPORT_FORMATS/
- Journal File Format: http://systemd.cn/docs/JOURNAL_FILE_FORMAT/
- Journal Message Catalogs: http://systemd.cn/docs/CATALOG/
- JSON Group Records: http://systemd.cn/docs/GROUP_RECORD/
- JSON User Records: http://systemd.cn/docs/USER_RECORD/
- Known Environment Variables: http://systemd.cn/docs/ENVIRONMENT/
- Locking Block Device Access: http://systemd.cn/docs/BLOCK_DEVICE_LOCKING/
- Minimal Builds: http://systemd.cn/docs/MINIMAL_BUILDS/
- My Service Can't Get Realtime!: http://systemd.cn/docs/MY_SERVICE_CANT_GET_REALTIME/
- Native Journal Protocol: http://systemd.cn/docs/JOURNAL_NATIVE_PROTOCOL/
- New Control Group Interfaces: http://systemd.cn/docs/CONTROL_GROUP_INTERFACE/
- Notes for Translators: http://systemd.cn/docs/TRANSLATORS/
- Password Agents: http://systemd.cn/docs/PASSWORD_AGENTS/
- Pax Controla Groupiana: http://systemd.cn/docs/PAX_CONTROL_GROUPS/
- Portability and Stability: http://systemd.cn/docs/PORTABILITY_AND_STABILITY/
- Portable Services Introduction: http://systemd.cn/docs/PORTABLE_SERVICES/
- Porting systemd To New Distributions: http://systemd.cn/docs/DISTRO_PORTING/
- Porting to New Architectures: http://systemd.cn/docs/PORTING_TO_NEW_ARCHITECTURES/
- Predictable Network Interface Names: http://systemd.cn/docs/PREDICTABLE_INTERFACE_NAMES/
- Presets: http://systemd.cn/docs/PRESET/
- Project IDs for Disk Quotas on Exec Directories: http://systemd.cn/docs/DISK-QUOTAS-PROJECTIDS/
- Random Seeds: http://systemd.cn/docs/RANDOM_SEEDS/
- Reporting of Security Vulnerabilities: http://systemd.cn/docs/SECURITY/
- Resource Pressure Handling: http://systemd.cn/docs/PRESSURE/
- Running Services After the Network Is Up: http://systemd.cn/docs/NETWORK_ONLINE/
- Safely Building Images: http://systemd.cn/docs/BUILDING_IMAGES/
- Socket Activation with Popular Daemons: http://systemd.cn/docs/DAEMON_SOCKET_ACTIVATION/
- Steps to a Successful Release: http://systemd.cn/docs/RELEASE/
- Storage Daemons for the Root File System: http://systemd.cn/docs/ROOT_STORAGE_DAEMONS/
- systemd Community Conduct Guidelines: http://systemd.cn/docs/CODE_OF_CONDUCT/
- systemd Coredump Handling: http://systemd.cn/docs/COREDUMP/
- systemd File Hierarchy Requirements: http://systemd.cn/docs/SYSTEMD_FILE_HIERARCHY_REQUIREMENTS/
- systemd Optimizations: http://systemd.cn/docs/OPTIMIZATIONS/
- systemd Repository Architecture: http://systemd.cn/docs/ARCHITECTURE/
- systemd-boot UEFI Boot Manager: http://systemd.cn/docs/BOOT/
- systemd-homed and JSON User/Group Record Support in Desktop Environments: http://systemd.cn/docs/USERDB_AND_DESKTOPS/
- systemd-resolved and VPNs: http://systemd.cn/docs/RESOLVED-VPNS/
- Testing systemd Using Sanitizers: http://systemd.cn/docs/TESTING_WITH_SANITIZERS/
- The Case for the /usr Merge: http://systemd.cn/docs/THE_CASE_FOR_THE_USR_MERGE/
- Tips And Tricks: http://systemd.cn/docs/TIPS_AND_TRICKS/
- User Record Blob Directories: http://systemd.cn/docs/USER_RECORD_BLOB_DIRS/
- User/Group Name Syntax: http://systemd.cn/docs/USER_NAMES/
- User/Group Record Lookup API via Varlink: http://systemd.cn/docs/USER_GROUP_API/
- Users, Groups, UIDs and GIDs on systemd Systems: http://systemd.cn/docs/UIDS-GIDS/
- Using /tmp/ and /var/tmp/ Safely: http://systemd.cn/docs/TEMPORARY_DIRECTORIES/
- Varlink API Style: http://systemd.cn/docs/VARLINK/
- VM Interface: http://systemd.cn/docs/VM_INTERFACE/
- What Settings Are Currently Available For Transient Units?: http://systemd.cn/docs/TRANSIENT-SETTINGS/
- Writing Desktop Environments: http://systemd.cn/docs/WRITING_DESKTOP_ENVIRONMENTS/
- Writing Display Managers: http://systemd.cn/docs/WRITING_DISPLAY_MANAGERS/
- Writing Network Configuration Managers: http://systemd.cn/docs/WRITING_NETWORK_CONFIGURATION_MANAGERS/
- Writing Resolver Clients: http://systemd.cn/docs/WRITING_RESOLVER_CLIENTS/
- Writing syslog Daemons Which Cooperate Nicely With systemd: http://systemd.cn/docs/SYSLOG/
- Writing VM and Container Managers: http://systemd.cn/docs/WRITING_VM_AND_CONTAINER_MANAGERS/


# My Service Can't Get Realtime!

_So, you have a service that requires real-time scheduling.
When you run this service on your systemd system it is unable to acquire real-time scheduling,
even though it is full root and has all possible privileges.
And now you are wondering what is going on and what you can do about it?_

## What is Going on?

By default systemd places all system services into their own control groups in the "cpu" hierarchy.
This has the benefit that the CPU usage of services with many worker threads or processes
(think: Apache with all its gazillion CGIs and stuff)
gets roughly the same amount of CPU as a service with very few worker threads (think: MySQL).
Instead of evening out CPU _per process_ this will cause CPU to be evened out _per service_.

Now, the "cpu" cgroup controller of the Linux kernel has one major shortcoming:
if a cgroup is created it needs an explicit, absolute RT time budget assigned,
or otherwise RT is not available to any process in the group, and an attempt to acquire it will fail with EPERM.
systemd will not assign any RT time budgets to the "cpu" cgroups it creates,
simply because there is no feasible way to do that,
since the budget needs to be specified in absolute time units and comes from a fixed pool.
Or in other words: we'd love to assign a budget, but there are no sane values we could use.
Thus, in its default configuration RT scheduling is simply not available for any system services.

## Working Around the Issue

Of course, that's quite a limitation, so here's how you work around this:

* One option is to simply globally turn off that systemd creates a "cpu" cgroup for each of the system services.
For that, edit `/etc/systemd/system.conf` and set `DefaultControllers=` to the empty string, then reboot.
(An alternative is to disable the "cpu" controller in your kernel, entirely.
systemd will not attempt to make use of controllers that aren't available in the kernel.)
* Another option is to turn this off for the specific service only.
For that, edit your service file, and add `ControlGroup=cpu:/` to its `[Service]` section.
This overrides the default logic for this one service only,
and places all its processes back in the root cgroup of the "cpu" hierarchy, which has the full RT budget assigned.
* A third option is to simply assign your service a realtime budget.
For that use `ControlGroupAttribute=cpu.rt_runtime_us 500000` in its `[Service]` or suchlike.
See [the kernel documentation](https://www.kernel.org/doc/Documentation/scheduler/sched-design-CFS.txt) for details.
The latter two options are not available for System V services.
A possible solution is to write a small wrapper service file that simply calls the SysV script's start verb in `ExecStart=` and the stop verb in `ExecStop=`.
(It also needs to set `RemainAfterExit=1` and `Type=forking`!)

Note that this all only applies to services.
By default, user applications run in the root cgroup of the "cpu" hierarchy, which avoids these problems for normal user applications.

In the long run we hope that the kernel is fixed to not require an RT budget to be assigned for any cgroup created before a process can acquire RT (i.e. a process' RT budget should be derived from the nearest ancestor cgroup which has a budget assigned, rather than unconditionally its own uninitialized budget.)
Ideally, we'd also like to create a per-user cgroup by default, so that users with many processes get roughly the same amount of CPU as users with very few.