diff --git a/Cargo.lock b/Cargo.lock
index f664af21..11683533 100644
--- a/Cargo.lock
+++ b/Cargo.lock
@@ -1922,7 +1922,6 @@ dependencies = [
  "binary-layout",
  "bincode",
  "criterion",
- "crossbeam-channel",
  "fast-counter",
  "hi_sparse_bitset",
  "human_bytes",
@@ -1959,7 +1958,6 @@ dependencies = [
  "bytes",
  "chrono",
  "chrono-tz",
- "crossbeam-channel",
  "decorum",
  "iana-time-zone",
  "inventory",
diff --git a/Cargo.toml b/Cargo.toml
index 0a295785..9810f833 100644
--- a/Cargo.toml
+++ b/Cargo.toml
@@ -59,7 +59,6 @@ binary-layout = "3.2.0"
 bytes = "1.5.0"
 chrono = "0.4.31"
 criterion = { version = "0.5.1", features = ["async_tokio"] }
-crossbeam-channel = "0.5.10"
 decorum = "0.3.1"  # For ordering & comparing our floats
 enum-primitive-derive = "0.3.0"
 fast-counter = "1.0.0"
diff --git a/crates/db/Cargo.toml b/crates/db/Cargo.toml
index efdc3166..4550b387 100644
--- a/crates/db/Cargo.toml
+++ b/crates/db/Cargo.toml
@@ -41,7 +41,6 @@ metrics.workspace = true
 metrics-macros.workspace = true
 
 # For the DB layer.
-crossbeam-channel.workspace = true
 bincode.workspace = true
 im.workspace = true
 sized-chunks.workspace = true
diff --git a/crates/db/src/tuplebox/coldstorage.rs b/crates/db/src/tuplebox/coldstorage.rs
index 86f3d885..5bb70854 100644
--- a/crates/db/src/tuplebox/coldstorage.rs
+++ b/crates/db/src/tuplebox/coldstorage.rs
@@ -272,7 +272,7 @@ impl ColdStorage {
                 Some(*r),
                 0,
                 ts,
-                page.page_size,
+                page.page_size as usize,
                 |buf| page.save_into(buf),
             );
             write_batch.push((*page_id, Some(wal_entry_buffer)));
diff --git a/crates/db/src/tuplebox/tb.rs b/crates/db/src/tuplebox/tb.rs
index 43dd4692..950a7602 100644
--- a/crates/db/src/tuplebox/tb.rs
+++ b/crates/db/src/tuplebox/tb.rs
@@ -15,7 +15,6 @@
 // TODO: support sorted indices, too.
 // TODO: 'join' and transitive closure -> datalog-style variable unification
 
-use std::collections::HashSet;
 use std::path::PathBuf;
 use std::sync::atomic::AtomicU64;
 use std::sync::Arc;
@@ -55,9 +54,6 @@ pub struct TupleBox {
     // TODO: take a look at Adnan's thread-sharded approach described in section 3.1
     //   (https://www.vldb.org/pvldb/vol16/p1426-alhomssi.pdf) -- "Ordered Snapshot Instant Commit"
     maximum_transaction: AtomicU64,
-    /// The set of currently active transactions, which will be used to prune old unreferenced
-    /// versions of tuples.
-    active_transactions: RwLock<HashSet<u64>>,
     /// Monotonically incrementing sequence counters.
     sequences: Vec<AtomicU64>,
     /// The copy-on-write set of current canonical base relations.
@@ -109,7 +105,6 @@ impl TupleBox {
         Arc::new(Self {
             relation_info: relations.to_vec(),
             maximum_transaction: AtomicU64::new(0),
-            active_transactions: RwLock::new(HashSet::new()),
             canonical: RwLock::new(base_relations),
             sequences,
             backing_store,
@@ -285,16 +280,10 @@ impl TupleBox {
             // And update the timestamp on the canonical relation.
             canonical[idx].ts = commit_ts;
         }
-        // Clear out the active transaction.
-        self.active_transactions.write().await.remove(&commit_ts);
 
         Ok(())
     }
 
-    pub(crate) async fn abort_transaction(&self, ts: u64) {
-        self.active_transactions.write().await.remove(&ts);
-    }
-
     pub async fn sync(&self, ts: u64, world_state: WorkingSet) {
         if let Some(bs) = &self.backing_store {
             let seqs = self
diff --git a/crates/db/src/tuplebox/tuples/slot_ptr.rs b/crates/db/src/tuplebox/tuples/slot_ptr.rs
index 5ddbe1cd..d943e3a1 100644
--- a/crates/db/src/tuplebox/tuples/slot_ptr.rs
+++ b/crates/db/src/tuplebox/tuples/slot_ptr.rs
@@ -26,7 +26,7 @@ use crate::tuplebox::tuples::{SlotBox, TupleId};
 pub struct SlotPtr {
     sb: Arc<SlotBox>,
     id: TupleId,
-    buflen: usize,
+    buflen: u32,
     bufaddr: *mut u8,
 
     _pin: std::marker::PhantomPinned,
@@ -46,7 +46,7 @@ impl SlotPtr {
             sb: sb.clone(),
             id: tuple_id,
             bufaddr,
-            buflen,
+            buflen: buflen as u32,
             _pin: std::marker::PhantomPinned,
         }
     }
@@ -85,7 +85,7 @@ impl SlotPtr {
     #[inline]
     fn buffer(&self) -> &[u8] {
         let buf_addr = self.as_ptr();
-        unsafe { std::slice::from_raw_parts(buf_addr, self.buflen) }
+        unsafe { std::slice::from_raw_parts(buf_addr, self.buflen as usize) }
     }
 
     #[inline]
diff --git a/crates/db/src/tuplebox/tuples/slotted_page.rs b/crates/db/src/tuplebox/tuples/slotted_page.rs
index db007260..99242f7c 100644
--- a/crates/db/src/tuplebox/tuples/slotted_page.rs
+++ b/crates/db/src/tuplebox/tuples/slotted_page.rs
@@ -37,7 +37,7 @@ use tracing::error;
 
 use crate::tuplebox::tuples::slotbox::SlotBoxError;
 
-pub type SlotId = usize;
+pub type SlotId = u32;
 
 // Note that if a page is empty, either because it's new, or because all its slots have been
 // removed, then used_bytes is 0.
@@ -173,7 +173,7 @@ impl SlotIndexEntry {
 /// is the same as the representation in-memory.
 pub struct SlottedPage<'a> {
     pub(crate) base_address: *mut u8,
-    pub(crate) page_size: usize,
+    pub(crate) page_size: u32,
 
     _marker: std::marker::PhantomData<&'a u8>,
 }
@@ -195,7 +195,7 @@ impl<'a> SlottedPage<'a> {
     pub fn for_page(base_address: *mut u8, page_size: usize) -> Self {
         Self {
             base_address,
-            page_size,
+            page_size: page_size as u32,
             _marker: Default::default(),
         }
     }
@@ -207,7 +207,7 @@ impl<'a> SlottedPage<'a> {
         let used = (header.num_slots * std::mem::size_of::<SlotIndexEntry>() as u32) as usize
             + header.used_bytes as usize
             + std::mem::size_of::<SlottedPageHeader>();
-        self.page_size - used
+        self.page_size as usize - used
     }
 
     /// How many bytes are available for appending to this page (i.e. not counting the space
@@ -218,7 +218,7 @@ impl<'a> SlottedPage<'a> {
         let index_length = header.index_length as usize;
         let header_size = std::mem::size_of::<SlottedPageHeader>();
 
-        self.page_size - (index_length + content_length + header_size)
+        self.page_size as usize - (index_length + content_length + header_size)
     }
 
     /// Add the slot into the page, copying it into the memory region, and returning the slot id
@@ -235,8 +235,9 @@ impl<'a> SlottedPage<'a> {
         }
         if let Some(fit_slot) = fit_slot {
             let content_position = self.offset_of(fit_slot).unwrap().0;
-            let memory_as_slice =
-                unsafe { std::slice::from_raw_parts_mut(self.base_address, self.page_size) };
+            let memory_as_slice = unsafe {
+                std::slice::from_raw_parts_mut(self.base_address, self.page_size as usize)
+            };
 
             // If there's an initial value provided, copy it in.
             if let Some(initial_value) = initial_value {
@@ -270,9 +271,9 @@ impl<'a> SlottedPage<'a> {
         // Add the slot to the content region. The start offset is PAGE_SIZE - content_length -
         // slot_length. So first thing, copy the bytes into the content region at that position.
         let content_length = self.header_mut().content_length as usize;
-        let content_position = self.page_size - content_length - size;
+        let content_position = self.page_size as usize - content_length - size;
         let memory_as_slice =
-            unsafe { std::slice::from_raw_parts_mut(self.base_address, self.page_size) };
+            unsafe { std::slice::from_raw_parts_mut(self.base_address, self.page_size as usize) };
 
         // If there's an initial value provided, copy it in.
         if let Some(initial_value) = initial_value {
@@ -282,8 +283,8 @@ impl<'a> SlottedPage<'a> {
         }
 
         // Add the index entry and expand the index region.
-        let mut index_entry = self.get_index_entry_mut(self.header_mut().num_slots as usize);
-        index_entry.as_mut().alloc(content_position, size);
+        let mut index_entry = self.get_index_entry_mut(self.header_mut().num_slots as SlotId);
+        index_entry.as_mut().alloc(content_position as usize, size);
 
         // Update the header
         let header = self.header_mut();
@@ -307,7 +308,7 @@ impl<'a> SlottedPage<'a> {
         let memory_as_slice = unsafe {
             Pin::new_unchecked(std::slice::from_raw_parts_mut(
                 self.base_address,
-                self.page_size,
+                self.page_size as usize,
             ))
         };
         lf(memory_as_slice);
@@ -321,7 +322,7 @@ impl<'a> SlottedPage<'a> {
         let mut slots = vec![];
         let num_slots = header.num_slots;
         for i in 0..num_slots {
-            let mut index_entry = self.get_index_entry_mut(i as usize);
+            let mut index_entry = self.get_index_entry_mut(i as SlotId);
             if index_entry.used {
                 unsafe { index_entry.as_mut().get_unchecked_mut() }.refcount = 1;
                 let slot_id = i as SlotId;
@@ -336,7 +337,7 @@ impl<'a> SlottedPage<'a> {
     pub fn save_into(&self, buf: &mut [u8]) {
         let _ = self.read_lock();
         let memory_as_slice =
-            unsafe { std::slice::from_raw_parts_mut(self.base_address, self.page_size) };
+            unsafe { std::slice::from_raw_parts_mut(self.base_address, self.page_size as usize) };
         buf.copy_from_slice(memory_as_slice);
     }
 
@@ -380,52 +381,52 @@ impl<'a> SlottedPage<'a> {
 
     fn get_slot(&self, slot_id: SlotId) -> Result<Pin<&'a [u8]>, SlotBoxError> {
         // Check that the index is in bounds
-        let num_slots = self.header().num_slots as usize;
+        let num_slots = self.header().num_slots as SlotId;
         if slot_id >= num_slots {
             error!(
                 "slot_id {} is out of bounds for page with {} slots",
                 slot_id, num_slots
             );
-            return Err(SlotBoxError::TupleNotFound(slot_id));
+            return Err(SlotBoxError::TupleNotFound(slot_id as usize));
         }
 
         // Read the index entry;
         let index_entry = self.get_index_entry(slot_id);
         if !index_entry.used {
             error!("slot_id {} is not used, invalid tuple", slot_id);
-            return Err(SlotBoxError::TupleNotFound(slot_id));
+            return Err(SlotBoxError::TupleNotFound(slot_id as usize));
         }
         let offset = index_entry.offset as usize;
         let length = index_entry.used_bytes as usize;
 
         let memory_as_slice =
-            unsafe { std::slice::from_raw_parts(self.base_address, self.page_size) };
+            unsafe { std::slice::from_raw_parts(self.base_address, self.page_size as usize) };
         Ok(unsafe { Pin::new_unchecked(&memory_as_slice[offset..offset + length]) })
     }
 
     fn get_slot_mut(&self, slot_id: SlotId) -> Result<Pin<&'a mut [u8]>, SlotBoxError> {
         // Check that the index is in bounds
-        let num_slots = self.header().num_slots as usize;
+        let num_slots = self.header().num_slots as SlotId;
         if slot_id >= num_slots {
-            return Err(SlotBoxError::TupleNotFound(slot_id));
+            return Err(SlotBoxError::TupleNotFound(slot_id as usize));
         }
 
         // Read the index entry;
         let index_entry = self.get_index_entry(slot_id);
         if !index_entry.used {
-            return Err(SlotBoxError::TupleNotFound(slot_id));
+            return Err(SlotBoxError::TupleNotFound(slot_id as usize));
         }
         let offset = index_entry.offset as usize;
         let length = index_entry.used_bytes as usize;
 
         assert!(
-            offset + length <= self.page_size,
+            offset + length <= self.page_size as usize,
             "slot {} is out of bounds",
             slot_id
         );
 
         let memory_as_slice =
-            unsafe { std::slice::from_raw_parts_mut(self.base_address, self.page_size) };
+            unsafe { std::slice::from_raw_parts_mut(self.base_address, self.page_size as usize) };
         Ok(unsafe { Pin::new_unchecked(&mut memory_as_slice[offset..offset + length]) })
     }
 }
@@ -525,9 +526,9 @@ impl<'a> SlottedPage<'a> {
     /// Return the offset, size of the slot at the given index.
     fn offset_of(&self, tid: SlotId) -> Result<(usize, usize), SlotBoxError> {
         // Check that the index is in bounds
-        let num_slots = self.header().num_slots as usize;
+        let num_slots = self.header().num_slots as SlotId;
         if tid >= num_slots {
-            return Err(SlotBoxError::TupleNotFound(tid));
+            return Err(SlotBoxError::TupleNotFound(tid as usize));
         }
 
         // Read the index entry;
@@ -541,7 +542,7 @@ impl<'a> SlottedPage<'a> {
         let header = self.header();
         let num_slots = header.num_slots;
         for i in 0..num_slots {
-            let index_entry = self.get_index_entry(i as usize);
+            let index_entry = self.get_index_entry(i as SlotId);
             if index_entry.used {
                 continue;
             }
@@ -554,7 +555,7 @@ impl<'a> SlottedPage<'a> {
         // Sort
         fits.sort_by(|a, b| a.1.cmp(&b.1));
         if let Some((tid, _)) = fits.first() {
-            return (true, Some(*tid));
+            return (true, Some(*tid as SlotId));
         }
 
         let index_length = header.index_length as isize;
@@ -569,7 +570,7 @@ impl<'a> SlottedPage<'a> {
 
     fn get_index_entry(&self, slot_id: SlotId) -> Pin<&SlotIndexEntry> {
         let index_offset = std::mem::size_of::<SlottedPageHeader>()
-            + (slot_id * std::mem::size_of::<SlotIndexEntry>());
+            + ((slot_id as usize) * std::mem::size_of::<SlotIndexEntry>());
 
         let base_address = self.base_address;
 
@@ -581,7 +582,7 @@ impl<'a> SlottedPage<'a> {
 
     fn get_index_entry_mut(&self, slot_id: SlotId) -> Pin<&mut SlotIndexEntry> {
         let index_offset = std::mem::size_of::<SlottedPageHeader>()
-            + (slot_id * std::mem::size_of::<SlotIndexEntry>());
+            + ((slot_id as usize) * std::mem::size_of::<SlotIndexEntry>());
         let base_address = self.base_address;
 
         unsafe {
@@ -593,7 +594,7 @@ impl<'a> SlottedPage<'a> {
 
 pub struct PageWriteGuard<'a> {
     base_address: *mut u8,
-    page_size: usize,
+    page_size: u32,
 
     _marker: std::marker::PhantomData<&'a u8>,
 }
@@ -657,7 +658,7 @@ impl<'a> Drop for PageWriteGuard<'a> {
 
 pub struct PageReadGuard<'a> {
     base_address: *const u8,
-    page_size: usize,
+    page_size: u32,
 
     _marker: std::marker::PhantomData<&'a u8>,
 }
@@ -673,7 +674,7 @@ impl<'a> PageReadGuard<'a> {
     pub fn get_slot(&self, slot_id: SlotId) -> Result<Pin<&'a [u8]>, SlotBoxError> {
         let sp = SlottedPage {
             base_address: self.base_address as _,
-            page_size: self.page_size,
+            page_size: self.page_size as _,
             _marker: Default::default(),
         };
         sp.get_slot(slot_id)
diff --git a/crates/db/src/tuplebox/tx/transaction.rs b/crates/db/src/tuplebox/tx/transaction.rs
index 44ff5f03..648af79d 100644
--- a/crates/db/src/tuplebox/tx/transaction.rs
+++ b/crates/db/src/tuplebox/tx/transaction.rs
@@ -31,8 +31,6 @@ use crate::tuplebox::RelationId;
 
 /// A versioned transaction, which is a fork of the current canonical base relations.
 pub struct Transaction {
-    /// The timestamp of this transaction, as granted to us by the tuplebox.
-    pub(crate) ts: u64,
     /// Where we came from, for referencing back to the base relations.
     db: Arc<TupleBox>,
     /// The "working set" is the set of retrieved and/or modified tuples from base relations, known
@@ -64,7 +62,6 @@ impl Transaction {
         let next_transient_relation_id = RelationId::transient(db.relation_info().len());
 
         Self {
-            ts,
             db,
             working_set: RwLock::new(Some(ws)),
             transient_relations: RwLock::new(HashMap::new()),
@@ -116,8 +113,6 @@ impl Transaction {
 
     pub async fn rollback(&self) -> Result<(), CommitError> {
         self.working_set.write().await.as_mut().unwrap().clear();
-        // Clear out the active transaction.
-        self.db.abort_transaction(self.ts).await;
         Ok(())
     }
 
diff --git a/crates/kernel/Cargo.toml b/crates/kernel/Cargo.toml
index 56c487ce..5518fdc3 100644
--- a/crates/kernel/Cargo.toml
+++ b/crates/kernel/Cargo.toml
@@ -61,5 +61,4 @@ metrics-util.workspace = true
 metrics-macros.workspace = true
 
 # For the DB layer.
-crossbeam-channel.workspace = true
 bincode.workspace = true