You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
* Fix a comment block disabling incorrectly RST form for cachalot_disabled
* Update CHANGELOG.rst
* Several points did I incorrectly say that 50 modifications per second was too much. It was supposed to say minute. You'd have to have a LARGE website to really get to that point since inserts from an external DB would typically take a 250-1000ms/1s per INSERT/UPDATE
* Add Python 3.9 support in the docs and setup.py
* Move chat to Discord and reflect in the docs
Cachalot officially supports Python 3.5-3.8 and Django 2.0-2.2, 3.0-3.1 with the databases PostgreSQL, SQLite, and MySQL.
42
+
Cachalot officially supports Python 3.5-3.9 and Django 2.0-2.2, 3.0-3.1 with the databases PostgreSQL, SQLite, and MySQL.
43
43
44
44
Usage
45
45
-----
@@ -99,7 +99,7 @@ Third-Party Cache Comparison
99
99
100
100
There are three main third party caches: cachalot, cache-machine, and cache-ops. Which do you use? We suggest a mix:
101
101
102
-
TL;DR Use cachalot for cold or modified <50 times per seconds (Most people should stick with only cachalot since you
102
+
TL;DR Use cachalot for cold or modified <50 times per minutes (Most people should stick with only cachalot since you
103
103
most likely won't need to scale to the point of needing cache-machine added to the bowl). If you're an enterprise that
104
104
already has huge statistics, then mixing cold caches for cachalot and your hot caches with cache-machine is the best
105
105
mix. However, when performing joins with ``select_related`` and ``prefetch_related``, you can
@@ -113,7 +113,7 @@ Cachalot is more-or-less intended for cold caches or "just-right" conditions. If
113
113
Django (also authored but work-in-progress by `Andrew Chen Wang`_), then the caching will work better since sharding
114
114
the cold/accessed-the-least records aren't invalidated as much.
115
115
116
-
Cachalot is good when there are <50 modifications per second on a hot cached table. This is mostly due to cache invalidation. It's the same with any cache,
116
+
Cachalot is good when there are <50 modifications per minute on a hot cached table. This is mostly due to cache invalidation. It's the same with any cache,
117
117
which is why we suggest you use cache-machine for hot caches. Cache-machine caches individual objects, taking up more in the memory store but
118
118
invalidates those individual objects instead of the entire table like cachalot.
@@ -64,7 +64,7 @@ In-depth opinion (from new maintainer):
64
64
65
65
There are three main third party caches: cachalot, cache-machine, and cache-ops. Which do you use? We suggest a mix:
66
66
67
-
TL;DR Use cachalot for cold or modified <50 times per seconds (Most people should stick with only cachalot since you
67
+
TL;DR Use cachalot for cold or modified <50 times per minutes (Most people should stick with only cachalot since you
68
68
most likely won't need to scale to the point of needing cache-machine added to the bowl). If you're an enterprise that
69
69
already has huge statistics, then mixing cold caches for cachalot and your hot caches with cache-machine is the best
70
70
mix. However, when performing joins with select_related and prefetch_related, you can
@@ -78,7 +78,7 @@ Cachalot is more-or-less intended for cold caches or "just-right" conditions. If
78
78
Django (also authored but work-in-progress by `Andrew Chen Wang <https://github.com/Andrew-Chen-Wang>`_),
79
79
then the caching will work better since sharding the cold/accessed-the-least records aren't invalidated as much.
80
80
81
-
Cachalot is good when there are <50 modifications per second on a hot cached table. This is mostly due to cache invalidation. It's the same with any cache,
81
+
Cachalot is good when there are <50 modifications per minute on a hot cached table. This is mostly due to cache invalidation. It's the same with any cache,
82
82
which is why we suggest you use cache-machine for hot caches. Cache-machine caches individual objects, taking up more in the memory store but
83
83
invalidates those individual objects instead of the entire table like cachalot.
0 commit comments