# Audited views configuration > **For AI agents:** the complete documentation index is at [llms.txt](https://questdb.com/docs/llms.txt). Every page is available as markdown by appending `.md` to its URL, or by sending an `Accept: text/markdown` request header. Configuration settings for audited views in QuestDB Enterprise. :::note Audited views are [Enterprise](/enterprise/) only. ::: An audited view records each read of it in the `sys.view_audit` table. These settings control how large a read's recorded parameters may be, the in-memory queue that carries rows from the reading query to the background job that writes them, and the storage policy the table is created with. None of these settings are reloadable: changing any of them requires a restart. The server checks the values at startup, and refuses to start, naming the property, when a numeric value is out of range. For example, to absorb longer bursts of audited reads and keep 90 days of the trail: ```ini title="server.conf" view.audit.queue.capacity=16384 view.audit.queue.params.max.size=256M view.audit.storage.policy=TO PARQUET 1d, DROP LOCAL 90d ``` For details, see [Audited views](/docs/security/audited-views/). ## view.audit.params.max.size - **Default**: `1M` (1 MiB) - **Reloadable**: no - **Range**: `1024` to `268435455` (1 KiB to 256 MiB minus one byte, the largest value a `VARCHAR` column holds) Largest `params` value, in bytes, that one read of an audited view may record. A read whose audited parameters render past it is refused with `view audit parameters are too large to record, the read is refused`, and records a row with status `error` and `NULL` in `params`. This keeps a caller from passing a value too large to record, such as a very long override of a list or text parameter. ## view.audit.queue.capacity - **Default**: `4096` - **Reloadable**: no - **Range**: `1` to `1073741824` (230) Number of audit rows the queue holds between the queries that read audited views and the job that writes them to `sys.view_audit`. The value is rounded up to a power of two, and the queue is allocated on the heap at startup. Recording never makes a read wait. When the queue is full, the read still runs, its row is dropped, and the server logs `view audit queue is full, dropping rows`. Raise the capacity if that message appears during bursts of audited reads. Auditing is lossy by design: see [Delivery guarantees](/docs/security/audited-views/#delivery) for every case in which a read goes unrecorded. ## view.audit.queue.params.max.size - **Default**: `64M` (64 MiB) - **Reloadable**: no - **Range**: at least [`view.audit.params.max.size`](#viewauditparamsmaxsize) Total size, in bytes, that the `params` of the rows waiting in the queue may add up to. It bounds the heap the queue can hold, which the slot count alone does not: each slot can hold up to [`view.audit.params.max.size`](#viewauditparamsmaxsize). A read whose row does not fit is still served, and its row is dropped as on a full queue, with the same log message. ## view.audit.storage.policy - **Default**: `TO PARQUET 1d` - **Reloadable**: no [Storage policy](/docs/concepts/storage-policy/) that `sys.view_audit` is created with. The value is the list of stages written inside `SET STORAGE POLICY(...)`, separated by commas, such as `TO PARQUET 1d, DROP LOCAL 90d`. The default converts each daily partition to Parquet one day after the partition ends, and never deletes a row. - The setting applies only when the server creates the table. Replicas take the table, and its policy, from the primary. - It does not change the policy of a table that already exists. Use [`ALTER TABLE SET STORAGE POLICY`](/docs/query/sql/alter-table-set-storage-policy/) for that. - Set the property to an empty value to create the table with no storage policy. If the server rejects the policy, it logs the error and creates the table without one. - A policy with `TO REMOTE` or `DROP REMOTE` is always rejected, because each node keeps its own trail under the same table name: see [Retention](/docs/security/audited-views/#retention).