Skip to content

mreg-api

Caching

Caching¶

The client can cache GET responses on disk to avoid refetching unchanged data. Caching is off by default. Any successful POST/PATCH/DELETE clears the cache, so writes never leave you reading stale data.

Enabling at construction¶

Pass cache=True for defaults, or a CacheConfig to customize:

from mreg_api import MregClient, CacheConfig

# Defaults
client = MregClient(url="https://mreg.example.com", cache=True)

# Custom configuration
client = MregClient(
    url="https://mreg.example.com",
    cache=CacheConfig(ttl=600, directory="/var/tmp/mreg-cache"),
)

`CacheConfig` options¶

Field	Type	Default	Description
`enable`	`bool`	`True`	Whether the cache is active.
`ttl`	`int`	`300`	Time-to-live for cached entries, in seconds.
`tag`	`str`	`"mreg-api"`	Tag used to namespace cache entries.
`timeout`	`int`	`60`	Open timeout for the underlying disk cache.
`directory`	`str \| None`	`None`	Cache directory. `None` uses a system temp directory.

Controlling the cache at runtime¶

client.enable_cache()                       # turn caching on
client.enable_cache(CacheConfig(ttl=60))    # reconfigure and turn on
client.disable_cache()                      # turn off (clears by default)
client.disable_cache(clear=False)           # turn off, keep cached data
client.clear_cache()                        # drop all cached entries

Use the caching() context manager to toggle caching for a block:

with client.caching(enable=True):
    client.host.list()  # served from / stored in the cache

Inspecting the cache¶

info = client.get_cache_info()  # CacheInfo, or None if caching is disabled
if info:
    print(info.items, info.hits, info.misses, info.size)