docs: Readability improvements (#4946)

Signed-off-by: Ryan Russell <git@ryanrussell.org>

Co-authored-by: kodiakhq[bot] <49736102+kodiakhq[bot]@users.noreply.github.com>

docs/ingester_querier_protocol.md

@@ -14,7 +14,7 @@ The `DoGet` ticket contains a [Protocol Buffer] message `influxdata.iox.ingester
   just ignore that column (i.e. the resulting data is the intersection of the request and the ingester data).
 - **predicate:** Predicate for row-filtering on the ingester side.
 
-The request does NOT contain a selection of partitions or sequencer. The ingester must respond with all partititions and
+The request does NOT contain a selection of partitions or sequencer. The ingester must respond with all partitions and
 sequencers it knows for that specified namespace-table combination.
 
 
@@ -67,7 +67,7 @@ Imagine the following ingester state:
   - partition P3:
     - no persisted parquet file
     - no max. persisted tombstone
-    - no snaphots (all deleted)
+    - no snapshots (all deleted)
   - partition P4:
     - no persisted parquet file
     - no max. persisted tombstone

docs/profiling.md

@@ -60,7 +60,7 @@ You will see a beautiful flame graph such as
 
 IOx includes a memory heap profile tool as well as a CPU profiler. The memory usage tool is based on [heappy](https://github.com/mkmik/heappy)
 
-Support is is not compiled in by defauly, but must be enabled via the `heappy` feature:
+Support is is not compiled in by default, but must be enabled via the `heappy` feature:
 
 ```shell
 # Compile and run IOx with heap profiling enabled

influxdb_iox_client/README.md

@@ -27,7 +27,7 @@ fn main() {
 
     let mut client = Client::new(connection);
 
-    // write a line of line procol data
+    // write a line of line protocol data
     client
         .write_lp("bananas", "cpu,region=west user=23.2 100",0)
         .await