From d018f2a0577172ccef8f28ca2e43f15be3e26616 Mon Sep 17 00:00:00 2001
From: roadscape <roadscape@users.noreply.github.com>
Date: Thu, 3 May 2018 15:29:02 -0500
Subject: [PATCH] bump up LRU post map to 2M entries
---
README.md | 2 +-
hive/indexer/posts.py | 3 ++-
2 files changed, 3 insertions(+), 2 deletions(-)
diff --git a/README.md b/README.md
index 0f91f2c77..f1464b8f6 100644
--- a/README.md
+++ b/README.md
@@ -68,7 +68,7 @@ Precedence: CLI over ENV over hive.conf. Check `hive --help` for details.
### Hardware
- Focus on Postgres performance
- - 2GB of memory for hive itself (TODO: verify/limit max usage during initial sync)
+ - 2.5GB of memory for `hive sync` process
- 200GB storage for database
diff --git a/hive/indexer/posts.py b/hive/indexer/posts.py
index f8f312dc4..58c3a7371 100644
--- a/hive/indexer/posts.py
+++ b/hive/indexer/posts.py
@@ -18,6 +18,7 @@ class Posts:
"""Handles critical/core post ops and data."""
# LRU cache for (author-permlink -> id) lookup
+ CACHE_SIZE = 2000000
_ids = collections.OrderedDict()
_hits = 0
_miss = 0
@@ -56,7 +57,7 @@ class Posts:
def _set_id(cls, url, pid):
"""Add an entry to the LRU, maintaining max size."""
assert pid, "no pid provided for %s" % url
- if len(cls._ids) > 1000000:
+ if len(cls._ids) > cls.CACHE_SIZE:
cls._ids.popitem(last=False)
cls._ids[url] = pid
--
GitLab