Comments on: MySQL: Enforcing long unique BLOB / TEXT columns with UNIQUE INDEX https://billauer.se/blog/2020/12/mysql-unique-index-blob-text/ Anything I found worthy to write down. Thu, 26 Mar 2026 13:15:15 +0000 hourly 1 http://wordpress.org/?v=3.1.2 By: eli https://billauer.se/blog/2020/12/mysql-unique-index-blob-text/comment-page-1/#comment-1560 eli Wed, 15 Dec 2021 09:03:25 +0000 https://billauer.se/blog/?p=6201#comment-1560 The question is how large system and what probability is acceptable. I Googled around a bit and found this: http://web.archive.org/web/20110725085124/http://bitcache.org/faq/hash-collision-probabilities If that's true, the probability for a hash collision is most likely negligible for any realistic database. The question is how large system and what probability is acceptable. I Googled around a bit and found this:

http://web.archive.org/web/20110725085124/http://bitcache.org/faq/hash-collision-probabilities

If that’s true, the probability for a hash collision is most likely negligible for any realistic database.

]]> By: Xiaotian Guo https://billauer.se/blog/2020/12/mysql-unique-index-blob-text/comment-page-1/#comment-1559 Xiaotian Guo Wed, 15 Dec 2021 01:08:43 +0000 https://billauer.se/blog/?p=6201#comment-1559 Eli, bumped into this post while researching similar topic - The way you implemented hash would not work actually. You will need to consider hash collision, otherwise your implementation will reject legit unique entries, although the chances are low, but it exists especially with a very large scale system. Eli, bumped into this post while researching similar topic – The way you implemented hash would not work actually. You will need to consider hash collision, otherwise your implementation will reject legit unique entries, although the chances are low, but it exists especially with a very large scale system.

]]>