diff --git a/site2/docs/reference-configuration.md b/site2/docs/reference-configuration.md
index 67ed800984d19..573a78f47508d 100644
--- a/site2/docs/reference-configuration.md
+++ b/site2/docs/reference-configuration.md
@@ -118,7 +118,7 @@ BookKeeper is a replicated log storage system that Pulsar uses for durable stora
 |bookieAuthProviderFactoryClass|The factory class name of the bookie authentication provider. If this is null, then there is no authentication.|null|
 |statsProviderClass||org.apache.bookkeeper.stats.prometheus.PrometheusMetricsProvider|
 |prometheusStatsHttpPort||8000|
-|dbStorage_writeCacheMaxSizeMb|Size of Write Cache. Memory is allocated from JVM direct memory. Write cache is used to buffer entries before flushing into the entry log. For good performance, it should be big enough to hold a substantial amount of entries in the flush interval. By default, it is allocated to 25% of the available direct memory.|N/A
+|dbStorage_writeCacheMaxSizeMb|Size of Write Cache. Memory is allocated from JVM direct memory. Write cache is used to buffer entries before flushing into the entry log. For good performance, it should be big enough to hold a substantial amount of entries in the flush interval.|25% of direct memory|
 |dbStorage_readAheadCacheMaxSizeMb|Size of Read cache. Memory is allocated from JVM direct memory. This read cache is pre-filled doing read-ahead whenever a cache miss happens. By default, it is allocated to 25% of the available direct memory.|N/A|
 |dbStorage_readAheadCacheBatchSize|How many entries to pre-fill in cache after a read cache miss|1000|
 |dbStorage_rocksDB_blockCacheSize|Size of RocksDB block-cache. For best performance, this cache should be big enough to hold a significant portion of the index database which can reach ~2GB in some cases. By default, it uses 10% of direct memory.|N/A|
diff --git a/site2/website/versioned_docs/version-2.1.0-incubating/reference-configuration.md b/site2/website/versioned_docs/version-2.1.0-incubating/reference-configuration.md
index 3d67d5f172010..e544b5a49ec58 100644
--- a/site2/website/versioned_docs/version-2.1.0-incubating/reference-configuration.md
+++ b/site2/website/versioned_docs/version-2.1.0-incubating/reference-configuration.md
@@ -85,7 +85,7 @@ BookKeeper is a replicated log storage system that Pulsar uses for durable stora
 |useHostNameAsBookieID|Whether the bookie should use its hostname to register with the coordination service (e.g.: zookeeper service). When false, bookie will use its ipaddress for the registration.|false|
 |statsProviderClass||org.apache.bookkeeper.stats.prometheus.PrometheusMetricsProvider|
 |prometheusStatsHttpPort||8000|
-|dbStorage_writeCacheMaxSizeMb|Size of Write Cache. Memory is allocated from JVM direct memory. Write cache is used to buffer entries before flushing into the entry log For good performance, it should be big enough to hold a sub|512|
+|dbStorage_writeCacheMaxSizeMb|Size of Write Cache. Memory is allocated from JVM direct memory. Write cache is used to buffer entries before flushing into the entry log For good performance, it should be big enough to hold a substantial amount of entries in the flush interval.|25% of direct memory|
 |dbStorage_readAheadCacheMaxSizeMb|Size of Read cache. Memory is allocated from JVM direct memory. This read cache is pre-filled doing read-ahead whenever a cache miss happens|256|
 |dbStorage_readAheadCacheBatchSize|How many entries to pre-fill in cache after a read cache miss|1000|
 |dbStorage_rocksDB_blockCacheSize|Size of RocksDB block-cache. For best performance, this cache should be big enough to hold a significant portion of the index database which can reach ~2GB in some cases|268435456|
diff --git a/site2/website/versioned_docs/version-2.1.1-incubating/reference-configuration.md b/site2/website/versioned_docs/version-2.1.1-incubating/reference-configuration.md
index ab8e387d11bb9..a2e0b5917710b 100644
--- a/site2/website/versioned_docs/version-2.1.1-incubating/reference-configuration.md
+++ b/site2/website/versioned_docs/version-2.1.1-incubating/reference-configuration.md
@@ -85,7 +85,7 @@ BookKeeper is a replicated log storage system that Pulsar uses for durable stora
 |useHostNameAsBookieID|Whether the bookie should use its hostname to register with the coordination service (e.g.: zookeeper service). When false, bookie will use its ipaddress for the registration.|false|
 |statsProviderClass||org.apache.bookkeeper.stats.prometheus.PrometheusMetricsProvider|
 |prometheusStatsHttpPort||8000|
-|dbStorage_writeCacheMaxSizeMb|Size of Write Cache. Memory is allocated from JVM direct memory. Write cache is used to buffer entries before flushing into the entry log For good performance, it should be big enough to hold a sub|512|
+|dbStorage_writeCacheMaxSizeMb|Size of Write Cache. Memory is allocated from JVM direct memory. Write cache is used to buffer entries before flushing into the entry log For good performance, it should be big enough to hold a substantial amount of entries in the flush interval.|25% of direct memory|
 |dbStorage_readAheadCacheMaxSizeMb|Size of Read cache. Memory is allocated from JVM direct memory. This read cache is pre-filled doing read-ahead whenever a cache miss happens|256|
 |dbStorage_readAheadCacheBatchSize|How many entries to pre-fill in cache after a read cache miss|1000|
 |dbStorage_rocksDB_blockCacheSize|Size of RocksDB block-cache. For best performance, this cache should be big enough to hold a significant portion of the index database which can reach ~2GB in some cases|268435456|
diff --git a/site2/website/versioned_docs/version-2.2.0/reference-configuration.md b/site2/website/versioned_docs/version-2.2.0/reference-configuration.md
index 10e76ea5d9a7e..27686489f5f45 100644
--- a/site2/website/versioned_docs/version-2.2.0/reference-configuration.md
+++ b/site2/website/versioned_docs/version-2.2.0/reference-configuration.md
@@ -85,7 +85,7 @@ BookKeeper is a replicated log storage system that Pulsar uses for durable stora
 |useHostNameAsBookieID|Whether the bookie should use its hostname to register with the coordination service (e.g.: zookeeper service). When false, bookie will use its ipaddress for the registration.|false|
 |statsProviderClass||org.apache.bookkeeper.stats.prometheus.PrometheusMetricsProvider|
 |prometheusStatsHttpPort||8000|
-|dbStorage_writeCacheMaxSizeMb|Size of Write Cache. Memory is allocated from JVM direct memory. Write cache is used to buffer entries before flushing into the entry log For good performance, it should be big enough to hold a sub|512|
+|dbStorage_writeCacheMaxSizeMb|Size of Write Cache. Memory is allocated from JVM direct memory. Write cache is used to buffer entries before flushing into the entry log For good performance, it should be big enough to hold a substantial amount of entries in the flush interval.|25% of direct memory|
 |dbStorage_readAheadCacheMaxSizeMb|Size of Read cache. Memory is allocated from JVM direct memory. This read cache is pre-filled doing read-ahead whenever a cache miss happens|256|
 |dbStorage_readAheadCacheBatchSize|How many entries to pre-fill in cache after a read cache miss|1000|
 |dbStorage_rocksDB_blockCacheSize|Size of RocksDB block-cache. For best performance, this cache should be big enough to hold a significant portion of the index database which can reach ~2GB in some cases|268435456|
diff --git a/site2/website/versioned_docs/version-2.2.1/reference-configuration.md b/site2/website/versioned_docs/version-2.2.1/reference-configuration.md
index 4ca50d4b5c2e3..09a81ff98bdd5 100644
--- a/site2/website/versioned_docs/version-2.2.1/reference-configuration.md
+++ b/site2/website/versioned_docs/version-2.2.1/reference-configuration.md
@@ -85,7 +85,7 @@ BookKeeper is a replicated log storage system that Pulsar uses for durable stora
 |useHostNameAsBookieID|Whether the bookie should use its hostname to register with the coordination service (e.g.: zookeeper service). When false, bookie will use its ipaddress for the registration.|false|
 |statsProviderClass||org.apache.bookkeeper.stats.prometheus.PrometheusMetricsProvider|
 |prometheusStatsHttpPort||8000|
-|dbStorage_writeCacheMaxSizeMb|Size of Write Cache. Memory is allocated from JVM direct memory. Write cache is used to buffer entries before flushing into the entry log For good performance, it should be big enough to hold a sub|512|
+|dbStorage_writeCacheMaxSizeMb|Size of Write Cache. Memory is allocated from JVM direct memory. Write cache is used to buffer entries before flushing into the entry log For good performance, it should be big enough to hold a substantial amount of entries in the flush interval.|25% of direct memory|
 |dbStorage_readAheadCacheMaxSizeMb|Size of Read cache. Memory is allocated from JVM direct memory. This read cache is pre-filled doing read-ahead whenever a cache miss happens|256|
 |dbStorage_readAheadCacheBatchSize|How many entries to pre-fill in cache after a read cache miss|1000|
 |dbStorage_rocksDB_blockCacheSize|Size of RocksDB block-cache. For best performance, this cache should be big enough to hold a significant portion of the index database which can reach ~2GB in some cases|268435456|
diff --git a/site2/website/versioned_docs/version-2.3.0/reference-configuration.md b/site2/website/versioned_docs/version-2.3.0/reference-configuration.md
index 8b765cd30a41b..c2a666869c316 100644
--- a/site2/website/versioned_docs/version-2.3.0/reference-configuration.md
+++ b/site2/website/versioned_docs/version-2.3.0/reference-configuration.md
@@ -85,7 +85,7 @@ BookKeeper is a replicated log storage system that Pulsar uses for durable stora
 |useHostNameAsBookieID|Whether the bookie should use its hostname to register with the coordination service (e.g.: zookeeper service). When false, bookie will use its ipaddress for the registration.|false|
 |statsProviderClass||org.apache.bookkeeper.stats.prometheus.PrometheusMetricsProvider|
 |prometheusStatsHttpPort||8000|
-|dbStorage_writeCacheMaxSizeMb|Size of Write Cache. Memory is allocated from JVM direct memory. Write cache is used to buffer entries before flushing into the entry log For good performance, it should be big enough to hold a sub|512|
+|dbStorage_writeCacheMaxSizeMb|Size of Write Cache. Memory is allocated from JVM direct memory. Write cache is used to buffer entries before flushing into the entry log For good performance, it should be big enough to hold a substantial amount of entries in the flush interval.|25% of direct memory|
 |dbStorage_readAheadCacheMaxSizeMb|Size of Read cache. Memory is allocated from JVM direct memory. This read cache is pre-filled doing read-ahead whenever a cache miss happens|256|
 |dbStorage_readAheadCacheBatchSize|How many entries to pre-fill in cache after a read cache miss|1000|
 |dbStorage_rocksDB_blockCacheSize|Size of RocksDB block-cache. For best performance, this cache should be big enough to hold a significant portion of the index database which can reach ~2GB in some cases|268435456|
diff --git a/site2/website/versioned_docs/version-2.3.1/reference-configuration.md b/site2/website/versioned_docs/version-2.3.1/reference-configuration.md
index 25f3a18fb5c65..b70866b720991 100644
--- a/site2/website/versioned_docs/version-2.3.1/reference-configuration.md
+++ b/site2/website/versioned_docs/version-2.3.1/reference-configuration.md
@@ -85,7 +85,7 @@ BookKeeper is a replicated log storage system that Pulsar uses for durable stora
 |useHostNameAsBookieID|Whether the bookie should use its hostname to register with the coordination service (e.g.: zookeeper service). When false, bookie will use its ipaddress for the registration.|false|
 |statsProviderClass||org.apache.bookkeeper.stats.prometheus.PrometheusMetricsProvider|
 |prometheusStatsHttpPort||8000|
-|dbStorage_writeCacheMaxSizeMb|Size of Write Cache. Memory is allocated from JVM direct memory. Write cache is used to buffer entries before flushing into the entry log For good performance, it should be big enough to hold a sub|512|
+|dbStorage_writeCacheMaxSizeMb|Size of Write Cache. Memory is allocated from JVM direct memory. Write cache is used to buffer entries before flushing into the entry log For good performance, it should be big enough to hold a substantial amount of entries in the flush interval.|25% of direct memory|
 |dbStorage_readAheadCacheMaxSizeMb|Size of Read cache. Memory is allocated from JVM direct memory. This read cache is pre-filled doing read-ahead whenever a cache miss happens|256|
 |dbStorage_readAheadCacheBatchSize|How many entries to pre-fill in cache after a read cache miss|1000|
 |dbStorage_rocksDB_blockCacheSize|Size of RocksDB block-cache. For best performance, this cache should be big enough to hold a significant portion of the index database which can reach ~2GB in some cases|268435456|
diff --git a/site2/website/versioned_docs/version-2.3.2/reference-configuration.md b/site2/website/versioned_docs/version-2.3.2/reference-configuration.md
index fe9149535eff8..1802b4ad8433b 100644
--- a/site2/website/versioned_docs/version-2.3.2/reference-configuration.md
+++ b/site2/website/versioned_docs/version-2.3.2/reference-configuration.md
@@ -85,7 +85,7 @@ BookKeeper is a replicated log storage system that Pulsar uses for durable stora
 |useHostNameAsBookieID|Whether the bookie should use its hostname to register with the coordination service (e.g.: zookeeper service). When false, bookie will use its ipaddress for the registration.|false|
 |statsProviderClass||org.apache.bookkeeper.stats.prometheus.PrometheusMetricsProvider|
 |prometheusStatsHttpPort||8000|
-|dbStorage_writeCacheMaxSizeMb|Size of Write Cache. Memory is allocated from JVM direct memory. Write cache is used to buffer entries before flushing into the entry log For good performance, it should be big enough to hold a sub|25% of direct memory|
+|dbStorage_writeCacheMaxSizeMb|Size of Write Cache. Memory is allocated from JVM direct memory. Write cache is used to buffer entries before flushing into the entry log For good performance, it should be big enough to hold a substantial amount of entries in the flush interval.|25% of direct memory|
 |dbStorage_readAheadCacheMaxSizeMb|Size of Read cache. Memory is allocated from JVM direct memory. This read cache is pre-filled doing read-ahead whenever a cache miss happens|25% of direct memory|
 |dbStorage_readAheadCacheBatchSize|How many entries to pre-fill in cache after a read cache miss|1000|
 |dbStorage_rocksDB_blockCacheSize|Size of RocksDB block-cache. For best performance, this cache should be big enough to hold a significant portion of the index database which can reach ~2GB in some cases|10% of direct memory|
diff --git a/site2/website/versioned_docs/version-2.4.0/reference-configuration.md b/site2/website/versioned_docs/version-2.4.0/reference-configuration.md
index 2e805d0399873..e7e72398d727e 100644
--- a/site2/website/versioned_docs/version-2.4.0/reference-configuration.md
+++ b/site2/website/versioned_docs/version-2.4.0/reference-configuration.md
@@ -87,7 +87,7 @@ BookKeeper is a replicated log storage system that Pulsar uses for durable stora
 |useHostNameAsBookieID|Whether the bookie should use its hostname to register with the coordination service (e.g.: zookeeper service). When false, bookie will use its ipaddress for the registration.|false|
 |statsProviderClass||org.apache.bookkeeper.stats.prometheus.PrometheusMetricsProvider|
 |prometheusStatsHttpPort||8000|
-|dbStorage_writeCacheMaxSizeMb|Size of Write Cache. Memory is allocated from JVM direct memory. Write cache is used to buffer entries before flushing into the entry log For good performance, it should be big enough to hold a sub|25% of direct memory|
+|dbStorage_writeCacheMaxSizeMb|Size of Write Cache. Memory is allocated from JVM direct memory. Write cache is used to buffer entries before flushing into the entry log For good performance, it should be big enough to hold a substantial amount of entries in the flush interval.|25% of direct memory|
 |dbStorage_readAheadCacheMaxSizeMb|Size of Read cache. Memory is allocated from JVM direct memory. This read cache is pre-filled doing read-ahead whenever a cache miss happens|25% of direct memory|
 |dbStorage_readAheadCacheBatchSize|How many entries to pre-fill in cache after a read cache miss|1000|
 |dbStorage_rocksDB_blockCacheSize|Size of RocksDB block-cache. For best performance, this cache should be big enough to hold a significant portion of the index database which can reach ~2GB in some cases|10% of direct memory|
diff --git a/site2/website/versioned_docs/version-2.4.1/reference-configuration.md b/site2/website/versioned_docs/version-2.4.1/reference-configuration.md
index 3fac3a084947a..e7c823b37acac 100644
--- a/site2/website/versioned_docs/version-2.4.1/reference-configuration.md
+++ b/site2/website/versioned_docs/version-2.4.1/reference-configuration.md
@@ -87,7 +87,7 @@ BookKeeper is a replicated log storage system that Pulsar uses for durable stora
 |useHostNameAsBookieID|Whether the bookie should use its hostname to register with the coordination service (e.g.: zookeeper service). When false, bookie will use its ipaddress for the registration.|false|
 |statsProviderClass||org.apache.bookkeeper.stats.prometheus.PrometheusMetricsProvider|
 |prometheusStatsHttpPort||8000|
-|dbStorage_writeCacheMaxSizeMb|Size of Write Cache. Memory is allocated from JVM direct memory. Write cache is used to buffer entries before flushing into the entry log For good performance, it should be big enough to hold a sub|25% of direct memory|
+|dbStorage_writeCacheMaxSizeMb|Size of Write Cache. Memory is allocated from JVM direct memory. Write cache is used to buffer entries before flushing into the entry log For good performance, it should be big enough to hold a substantial amount of entries in the flush interval.|25% of direct memory|
 |dbStorage_readAheadCacheMaxSizeMb|Size of Read cache. Memory is allocated from JVM direct memory. This read cache is pre-filled doing read-ahead whenever a cache miss happens|25% of direct memory|
 |dbStorage_readAheadCacheBatchSize|How many entries to pre-fill in cache after a read cache miss|1000|
 |dbStorage_rocksDB_blockCacheSize|Size of RocksDB block-cache. For best performance, this cache should be big enough to hold a significant portion of the index database which can reach ~2GB in some cases|10% of direct memory|
diff --git a/site2/website/versioned_docs/version-2.4.2/reference-configuration.md b/site2/website/versioned_docs/version-2.4.2/reference-configuration.md
index 375b6de954a84..890181abd8b73 100644
--- a/site2/website/versioned_docs/version-2.4.2/reference-configuration.md
+++ b/site2/website/versioned_docs/version-2.4.2/reference-configuration.md
@@ -87,7 +87,7 @@ BookKeeper is a replicated log storage system that Pulsar uses for durable stora
 |useHostNameAsBookieID|Whether the bookie should use its hostname to register with the coordination service (e.g.: zookeeper service). When false, bookie will use its ipaddress for the registration.|false|
 |statsProviderClass||org.apache.bookkeeper.stats.prometheus.PrometheusMetricsProvider|
 |prometheusStatsHttpPort||8000|
-|dbStorage_writeCacheMaxSizeMb|Size of Write Cache. Memory is allocated from JVM direct memory. Write cache is used to buffer entries before flushing into the entry log For good performance, it should be big enough to hold a sub|25% of direct memory|
+|dbStorage_writeCacheMaxSizeMb|Size of Write Cache. Memory is allocated from JVM direct memory. Write cache is used to buffer entries before flushing into the entry log For good performance, it should be big enough to hold a substantial amount of entries in the flush interval.|25% of direct memory|
 |dbStorage_readAheadCacheMaxSizeMb|Size of Read cache. Memory is allocated from JVM direct memory. This read cache is pre-filled doing read-ahead whenever a cache miss happens|25% of direct memory|
 |dbStorage_readAheadCacheBatchSize|How many entries to pre-fill in cache after a read cache miss|1000|
 |dbStorage_rocksDB_blockCacheSize|Size of RocksDB block-cache. For best performance, this cache should be big enough to hold a significant portion of the index database which can reach ~2GB in some cases|10% of direct memory|
diff --git a/site2/website/versioned_docs/version-2.5.0/reference-configuration.md b/site2/website/versioned_docs/version-2.5.0/reference-configuration.md
index 365aa43497d92..dc1ad2966ea49 100644
--- a/site2/website/versioned_docs/version-2.5.0/reference-configuration.md
+++ b/site2/website/versioned_docs/version-2.5.0/reference-configuration.md
@@ -86,7 +86,7 @@ BookKeeper is a replicated log storage system that Pulsar uses for durable stora
 |useHostNameAsBookieID|Whether the bookie should use its hostname to register with the coordination service (e.g.: zookeeper service). When false, bookie will use its ipaddress for the registration.|false|
 |statsProviderClass||org.apache.bookkeeper.stats.prometheus.PrometheusMetricsProvider|
 |prometheusStatsHttpPort||8000|
-|dbStorage_writeCacheMaxSizeMb|Size of Write Cache. Memory is allocated from JVM direct memory. Write cache is used to buffer entries before flushing into the entry log For good performance, it should be big enough to hold a sub|25% of direct memory|
+|dbStorage_writeCacheMaxSizeMb|Size of Write Cache. Memory is allocated from JVM direct memory. Write cache is used to buffer entries before flushing into the entry log For good performance, it should be big enough to hold a substantial amount of entries in the flush interval.|25% of direct memory|
 |dbStorage_readAheadCacheMaxSizeMb|Size of Read cache. Memory is allocated from JVM direct memory. This read cache is pre-filled doing read-ahead whenever a cache miss happens|25% of direct memory|
 |dbStorage_readAheadCacheBatchSize|How many entries to pre-fill in cache after a read cache miss|1000|
 |dbStorage_rocksDB_blockCacheSize|Size of RocksDB block-cache. For best performance, this cache should be big enough to hold a significant portion of the index database which can reach ~2GB in some cases|10% of direct memory|
diff --git a/site2/website/versioned_docs/version-2.5.1/reference-configuration.md b/site2/website/versioned_docs/version-2.5.1/reference-configuration.md
index ff4a941406471..1a007882c36dc 100644
--- a/site2/website/versioned_docs/version-2.5.1/reference-configuration.md
+++ b/site2/website/versioned_docs/version-2.5.1/reference-configuration.md
@@ -86,7 +86,7 @@ BookKeeper is a replicated log storage system that Pulsar uses for durable stora
 |useHostNameAsBookieID|Whether the bookie should use its hostname to register with the coordination service (e.g.: zookeeper service). When false, bookie will use its ipaddress for the registration.|false|
 |statsProviderClass||org.apache.bookkeeper.stats.prometheus.PrometheusMetricsProvider|
 |prometheusStatsHttpPort||8000|
-|dbStorage_writeCacheMaxSizeMb|Size of Write Cache. Memory is allocated from JVM direct memory. Write cache is used to buffer entries before flushing into the entry log For good performance, it should be big enough to hold a sub|25% of direct memory|
+|dbStorage_writeCacheMaxSizeMb|Size of Write Cache. Memory is allocated from JVM direct memory. Write cache is used to buffer entries before flushing into the entry log For good performance, it should be big enough to hold a substantial amount of entries in the flush interval.|25% of direct memory|
 |dbStorage_readAheadCacheMaxSizeMb|Size of Read cache. Memory is allocated from JVM direct memory. This read cache is pre-filled doing read-ahead whenever a cache miss happens|25% of direct memory|
 |dbStorage_readAheadCacheBatchSize|How many entries to pre-fill in cache after a read cache miss|1000|
 |dbStorage_rocksDB_blockCacheSize|Size of RocksDB block-cache. For best performance, this cache should be big enough to hold a significant portion of the index database which can reach ~2GB in some cases|10% of direct memory|
diff --git a/site2/website/versioned_docs/version-2.5.2/reference-configuration.md b/site2/website/versioned_docs/version-2.5.2/reference-configuration.md
index 362988c3d6611..7b63a0bd823ce 100644
--- a/site2/website/versioned_docs/version-2.5.2/reference-configuration.md
+++ b/site2/website/versioned_docs/version-2.5.2/reference-configuration.md
@@ -87,7 +87,7 @@ BookKeeper is a replicated log storage system that Pulsar uses for durable stora
 |useHostNameAsBookieID|Whether the bookie should use its hostname to register with the coordination service (e.g.: zookeeper service). When false, bookie will use its ipaddress for the registration.|false|
 |statsProviderClass||org.apache.bookkeeper.stats.prometheus.PrometheusMetricsProvider|
 |prometheusStatsHttpPort||8000|
-|dbStorage_writeCacheMaxSizeMb|Size of Write Cache. Memory is allocated from JVM direct memory. Write cache is used to buffer entries before flushing into the entry log For good performance, it should be big enough to hold a sub|25% of direct memory|
+|dbStorage_writeCacheMaxSizeMb|Size of Write Cache. Memory is allocated from JVM direct memory. Write cache is used to buffer entries before flushing into the entry log For good performance, it should be big enough to hold a substantial amount of entries in the flush interval.|25% of direct memory|
 |dbStorage_readAheadCacheMaxSizeMb|Size of Read cache. Memory is allocated from JVM direct memory. This read cache is pre-filled doing read-ahead whenever a cache miss happens|25% of direct memory|
 |dbStorage_readAheadCacheBatchSize|How many entries to pre-fill in cache after a read cache miss|1000|
 |dbStorage_rocksDB_blockCacheSize|Size of RocksDB block-cache. For best performance, this cache should be big enough to hold a significant portion of the index database which can reach ~2GB in some cases|10% of direct memory|
diff --git a/site2/website/versioned_docs/version-2.6.0/reference-configuration.md b/site2/website/versioned_docs/version-2.6.0/reference-configuration.md
index 65fd1ff932f19..12157717bf344 100644
--- a/site2/website/versioned_docs/version-2.6.0/reference-configuration.md
+++ b/site2/website/versioned_docs/version-2.6.0/reference-configuration.md
@@ -87,7 +87,7 @@ BookKeeper is a replicated log storage system that Pulsar uses for durable stora
 |useHostNameAsBookieID|Whether the bookie should use its hostname to register with the coordination service (e.g.: zookeeper service). When false, bookie will use its ipaddress for the registration.|false|
 |statsProviderClass||org.apache.bookkeeper.stats.prometheus.PrometheusMetricsProvider|
 |prometheusStatsHttpPort||8000|
-|dbStorage_writeCacheMaxSizeMb|Size of Write Cache. Memory is allocated from JVM direct memory. Write cache is used to buffer entries before flushing into the entry log For good performance, it should be big enough to hold a sub|25% of direct memory|
+|dbStorage_writeCacheMaxSizeMb|Size of Write Cache. Memory is allocated from JVM direct memory. Write cache is used to buffer entries before flushing into the entry log For good performance, it should be big enough to hold a substantial amount of entries in the flush interval.|25% of direct memory|
 |dbStorage_readAheadCacheMaxSizeMb|Size of Read cache. Memory is allocated from JVM direct memory. This read cache is pre-filled doing read-ahead whenever a cache miss happens|25% of direct memory|
 |dbStorage_readAheadCacheBatchSize|How many entries to pre-fill in cache after a read cache miss|1000|
 |dbStorage_rocksDB_blockCacheSize|Size of RocksDB block-cache. For best performance, this cache should be big enough to hold a significant portion of the index database which can reach ~2GB in some cases|10% of direct memory|
diff --git a/site2/website/versioned_docs/version-2.6.1/reference-configuration.md b/site2/website/versioned_docs/version-2.6.1/reference-configuration.md
index c0179c87de504..46dd8aa102f11 100644
--- a/site2/website/versioned_docs/version-2.6.1/reference-configuration.md
+++ b/site2/website/versioned_docs/version-2.6.1/reference-configuration.md
@@ -87,7 +87,7 @@ BookKeeper is a replicated log storage system that Pulsar uses for durable stora
 |useHostNameAsBookieID|Whether the bookie should use its hostname to register with the coordination service (e.g.: zookeeper service). When false, bookie will use its ipaddress for the registration.|false|
 |statsProviderClass||org.apache.bookkeeper.stats.prometheus.PrometheusMetricsProvider|
 |prometheusStatsHttpPort||8000|
-|dbStorage_writeCacheMaxSizeMb|Size of Write Cache. Memory is allocated from JVM direct memory. Write cache is used to buffer entries before flushing into the entry log For good performance, it should be big enough to hold a sub|25% of direct memory|
+|dbStorage_writeCacheMaxSizeMb|Size of Write Cache. Memory is allocated from JVM direct memory. Write cache is used to buffer entries before flushing into the entry log For good performance, it should be big enough to hold a substantial amount of entries in the flush interval.|25% of direct memory|
 |dbStorage_readAheadCacheMaxSizeMb|Size of Read cache. Memory is allocated from JVM direct memory. This read cache is pre-filled doing read-ahead whenever a cache miss happens|25% of direct memory|
 |dbStorage_readAheadCacheBatchSize|How many entries to pre-fill in cache after a read cache miss|1000|
 |dbStorage_rocksDB_blockCacheSize|Size of RocksDB block-cache. For best performance, this cache should be big enough to hold a significant portion of the index database which can reach ~2GB in some cases|10% of direct memory|
diff --git a/site2/website/versioned_docs/version-2.6.2/reference-configuration.md b/site2/website/versioned_docs/version-2.6.2/reference-configuration.md
index 32ec85392a5f8..0be83e0d2a32e 100644
--- a/site2/website/versioned_docs/version-2.6.2/reference-configuration.md
+++ b/site2/website/versioned_docs/version-2.6.2/reference-configuration.md
@@ -86,7 +86,7 @@ BookKeeper is a replicated log storage system that Pulsar uses for durable stora
 |useHostNameAsBookieID|Whether the bookie should use its hostname to register with the coordination service (e.g.: zookeeper service). When false, bookie will use its ipaddress for the registration.|false|
 |statsProviderClass||org.apache.bookkeeper.stats.prometheus.PrometheusMetricsProvider|
 |prometheusStatsHttpPort||8000|
-|dbStorage_writeCacheMaxSizeMb|Size of Write Cache. Memory is allocated from JVM direct memory. Write cache is used to buffer entries before flushing into the entry log For good performance, it should be big enough to hold a sub|25% of direct memory|
+|dbStorage_writeCacheMaxSizeMb|Size of Write Cache. Memory is allocated from JVM direct memory. Write cache is used to buffer entries before flushing into the entry log For good performance, it should be big enough to hold substantial amount of entries in the flush interval.|25% of direct memory|
 |dbStorage_readAheadCacheMaxSizeMb|Size of Read cache. Memory is allocated from JVM direct memory. This read cache is pre-filled doing read-ahead whenever a cache miss happens|25% of direct memory|
 |dbStorage_readAheadCacheBatchSize|How many entries to pre-fill in cache after a read cache miss|1000|
 |dbStorage_rocksDB_blockCacheSize|Size of RocksDB block-cache. For best performance, this cache should be big enough to hold a significant portion of the index database which can reach ~2GB in some cases|10% of direct memory|