```{index} single: hash_seed; Hash::Util function ``` ```{index} single: Hash::Util::hash_seed; Perl function ``` # hash_seed Return the per-process random bytes that seed the hash function. ## Synopsis my $seed = hash_seed(); The result is a byte string whose length depends on the hash algorithm built into the interpreter — commonly `4` bytes for the classic algorithms and `16` bytes for SipHash. Two interpreters started from the same program see different values; a single run sees the same value throughout its life. The seed is sensitive. Anyone who learns it can construct keys that collide in your hashes and mount an algorithmic-complexity attack. Keep it out of logs and error messages.