Make redis connection pool max sizes configurable, bump default

PR-URL: hasura/graphql-engine-mono#11382
GitOrigin-RevId: becf8cf3a07ed1fcc824b479c056d04265058817

Author: jberryman

docs/docs/caching/enterprise-caching.mdx

@@ -66,12 +66,14 @@ HASURA_GRAPHQL_RATE_LIMIT_REDIS_TLS_HOSTNAME="redishostname"
 You can tune the various caching parameters according to your use-case. For example, you can configure the maximum
 TTL allowed, the max entry size in the cache etc.
 
-| Env var                               | Flag                           | Description                                                                                                                                                  |
-|---------------------------------------|--------------------------------|--------------------------------------------------------------------------------------------------------------------------------------------------------------|
-| `HASURA_GRAPHQL_CACHE_MAX_ENTRY_TTL`  | `--query-cache-max-ttl`        | Maximum TTL allowed in seconds. Clients can request TTL via the `@cached` directive. But an upper limit can be set using this setting. Default: 3600 seconds |
-| `HASURA_GRAPHQL_CACHE_MAX_ENTRY_SIZE` | `--query-cache-max-entry-size` | Maximum size of the response that is allowed to be cached (in MB). default: 1000 MB                                                                          |
-| `HASURA_GRAPHQL_CACHE_BUCKET_RATE`    | `--query-cache-bucket-rate`    | Recharge rate for the Query Response Cache token bucket. Default: 10,000,000 bytes/second (10 MB/s)                                                          |
-| `HASURA_GRAPHQL_CACHE_BUCKET_SIZE`    | `--query-cache-bucket-size`    | Maximum capacity in bytes for the Query Response Cache token bucket algorithm. See https://hasura.io/docs/latest/queries/response-caching for more info. Default: 1000000000 bytes (1 GB) |
+| Env var                                        | Flag                                 | Description                                                                                                                                                  |
+|------------------------------------------------|--------------------------------------|--------------------------------------------------------------------------------------------------------------------------------------------------------------|
+| `HASURA_GRAPHQL_CACHE_MAX_ENTRY_TTL`           | `--query-cache-max-ttl`              | Maximum TTL allowed in seconds. Clients can request TTL via the `@cached` directive. But an upper limit can be set using this setting. Default: 3600 seconds |
+| `HASURA_GRAPHQL_CACHE_MAX_ENTRY_SIZE`          | `--query-cache-max-entry-size`       | Maximum size of the response that is allowed to be cached (in MB). default: 1000 MB                                                                          |
+| `HASURA_GRAPHQL_CACHE_BUCKET_RATE`             | `--query-cache-bucket-rate`          | Recharge rate for the Query Response Cache token bucket. Default: 10,000,000 bytes/second (10 MB/s)                                                          |
+| `HASURA_GRAPHQL_CACHE_BUCKET_SIZE`             | `--query-cache-bucket-size`          | Maximum capacity in bytes for the Query Response Cache token bucket algorithm. See https://hasura.io/docs/latest/queries/response-caching for more info. Default: 1000000000 bytes (1 GB) |
+| `HASURA_GRAPHQL_CACHING_REDIS_POOL_SIZE`       | `--caching-redis-pool-size`          | Maximum number of connections in the caching Redis connection pool. Default: 500                                                                             |
+| `HASURA_GRAPHQL_RATE_LIMIT_REDIS_POOL_SIZE`    | `--rate-limit-redis-pool-size`       | Maximum number of connections in the rate-limiting Redis connection pool. Default: 500                                                                       |
 
 
 ### How to tune these parameters

docs/docs/deployment/graphql-engine-flags/reference.mdx

@@ -1101,6 +1101,30 @@ The path to a shared CA store to use to connect to both (caching and rate-limiti
 | **Default**         | `null`                                                       |
 | **Supported in**    | Enterprise Edition only                                      |
 
+### Caching Redis Pool Size
+
+Maximum number of connections in the caching Redis connection pool. For high-traffic production environments, you may want to increase this value. For development or low-traffic environments, you can decrease it to conserve resources.
+
+|                     |                                             |
+| ------------------- | ------------------------------------------- |
+| **Flag**            | `--caching-redis-pool-size`                 |
+| **Env var**         | `HASURA_GRAPHQL_CACHING_REDIS_POOL_SIZE`    |
+| **Accepted values** | Integer                                     |
+| **Default**         | `500`                                       |
+| **Supported in**    | Enterprise Edition only                     |
+
+### Rate Limit Redis Pool Size
+
+Maximum number of connections in the rate-limiting Redis connection pool. For high-traffic production environments with heavy rate-limiting usage, you may want to increase this value. For development or low-traffic environments, you can decrease it to conserve resources.
+
+|                     |                                             |
+| ------------------- | ------------------------------------------- |
+| **Flag**            | `--rate-limit-redis-pool-size`              |
+| **Env var**         | `HASURA_GRAPHQL_RATE_LIMIT_REDIS_POOL_SIZE` |
+| **Accepted values** | Integer                                     |
+| **Default**         | `500`                                       |
+| **Supported in**    | Enterprise Edition only                     |
+
 ### Redis URL
 
 The Redis URL to use for [query caching](/caching/enterprise-caching.mdx) and

server/forks/hedis/src/Database/Redis/Connection.hs

@@ -107,7 +107,7 @@ instance Exception ConnectError
 --  connectPort           = PortNumber 6379 -- Redis default port
 --  connectAuth           = Nothing         -- No password
 --  connectDatabase       = 0               -- SELECT database 0
---  connectMaxConnections = 50              -- Up to 50 connections
+--  connectMaxConnections = 500             -- Up to 500 connections
 --  connectMaxIdleTime    = 30              -- Keep open for 30 seconds
 --  connectTimeout        = Nothing         -- Don't add timeout logic
 --  connectTLSParams      = Nothing         -- Do not use TLS
@@ -120,7 +120,10 @@ defaultConnectInfo = ConnInfo
     , connectAuth           = Nothing
     , connectReadOnly       = False
     , connectDatabase       = 0
-    , connectMaxConnections = 50
+    -- This keeps us several orders of magnitude below default limit, and
+    -- typical limits of managed services AFAICT.  Also see:
+    -- https://redis.io/docs/latest/operate/oss_and_stack/management/optimization/benchmarks/?utm_source=chatgpt.com#factors-impacting-redis-performance
+    , connectMaxConnections = 500
     , connectMaxIdleTime    = 30
     , connectTimeout        = Nothing
     , connectTLSParams      = Nothing