Final answer:
Tokens in Splunk are stored in the lexicon and the associated posting lists post-indexing, which enables efficient search and retrieval of data.
Step-by-step explanation:
After Splunk tokenizes terms at index time, the tokens are stored in the lexicon and posting lists. The lexicon is a structure which stores all the unique words encountered during indexing, which are then associated with a list of all the events that contain them, known as the posting list. Each indexed event in Splunk is assigned an event ID, and the posting list uses these IDs to reference events that contain the particular term. This mechanism allows for efficient searching and retrieval of indexed data when a search query is entered in Splunk.
It is not stored in a bloom filter, as bloom filters are used to test whether an element is a member of a set and are not used for storing tokens. A frozen bucket is a term in Splunk related to data lifecycle management, representing a stage where data is archived and is no longer searchable, so tokens are not stored there directly either.