3x Faster TID Range Scans - Postgres 19

If you've ever had to scrub significantly large tables—whether updating older records or deleting expired ones—you know the pain of trying to do it efficiently without locking the entire table. One common trick is to iterate over the ctid (the physical location of the tuple) to process the table in chunks. Until now, this was strictly a single-threaded affair. But with the latest commit to Postgres 19, TID Range Scans can now be parallelized, allowing you to harness multiple cores to rip through these maintenance tasks significantly faster.

In my tests, upto 3x Faster!!

The Problem with Linear Scanning

When you perform a batched operation using ctid (e.g., WHERE ctid >= '(0,0)' AND ctid < '(1000,0)'), Postgres uses a Tid Range Scan. Previously, this scan type didn't support parallelism. If you had a 1 TB table and wanted to process ranges of it, a single CPU core would have to do all the heavy lifting for that range, fetching blocks one by one. Even if you had 64 cores gathering dust, they couldn't help with that specific scan node.

The Solution: Parallel TID Range Scan

Commit 0ca3b169, developed by Cary Huang and committed by David Rowley, introduces infrastructure to allow Tid Range Scan to participate in parallel query plans. The logic effectively splits the block range among the available parallel workers. Instead of one process sweeping from block 0 to N, workers can grab chunks of blocks concurrently.

I was following the discussion on the mailing list, where Cary and David collaborated closely to refine the parallel safety logic, ensuring workers handle scan limits correctly—leading to the robust implementation that landed in the master branch.

NOTE

This feature is currently in the master branch (Postgres 19devel). As with all unreleased features, it could theoretically be rolled back or modified before the final release.

Benchmark: Before vs. After

To see this in action, I spun up two Postgres instances: one built just before the commit (Postgres 18) and one after (Postgres 19). I created a table bench_tid_range with 10 million rows and ran a count(*) query over the first 50% of the table using a ctid range condition.

Test Setup:

• Table Size: 10 Million rows
• Query: SELECT count(*) FROM bench_tid_range WHERE ctid >= '(0,0)' AND ctid < '(41667,0)'

Environment	Workers	Execution Time (Median)	Speedup
Before (Pg 18)	0	448 ms	1.00x
After (Pg 19)	0	435 ms	1.03x
After (Pg 19)	1	238 ms	1.88x
After (Pg 19)	2	174 ms	2.58x
After (Pg 19)	3	151 ms	2.97x
After (Pg 19)	4	150 ms	2.98x
After (Pg 19)	5	147 ms	3.05x
After (Pg 19)	6	143 ms	3.14x
After (Pg 19)	7	147 ms	3.04x
After (Pg 19)	8	147 ms	3.04x

*(Each execution time is a Median of 9 runs)*

The chart above illustrates the performance scaling. Note that Postgres 19 with 0 workers (i.e. serial) is almost identical in speed to Postgres 18; the efficiency gains are visible only when parallelism is enabled.

We see a massive drop in execution time just by enabling 1 worker (which effectively gives us 2 processes scanning: the leader + 1 worker). The gains continue robustly up to 3 workers, where execution time settles around 150ms. Beyond that, we hit diminishing returns as the overhead of managing parallel workers and aggregating results begins to dominate the raw scanning speed. The "sweet spot" for this specific workload appears to be around 2-3 workers.

Under the Hood: Deep Dive

Let's look at the actual EXPLAIN (ANALYZE, BUFFERS) output to see *why* it's faster.

Before (Postgres 18)

The legacy behavior forces a serial scan. The single process reads all 41,667 pages.

 Aggregate  (cost=104171.87..104171.88 rows=1 width=8) (actual time=427.723..427.724 rows=1.00 loops=1)
   Buffers: shared read=41667
   ->  Tid Range Scan on bench_tid_range  (cost=0.01..91671.70 rows=5000069 width=0) (actual time=0.130..243.502 rows=5000040.00 loops=1)
         TID Cond: ((ctid >= '(0,0)'::tid) AND (ctid < '(41667,0)'::tid))
         Buffers: shared read=41667

After (Postgres 19)

The new Parallel Tid Range Scan node appears. Notice the Gather node launching 2 workers.

 Finalize Aggregate  (cost=68713.25..68713.26 rows=1 width=8) (actual time=166.873..169.987 rows=1.00 loops=1)
   Buffers: shared read=41667
   ->  Gather  (cost=68713.03..68713.24 rows=2 width=8) (actual time=166.238..169.978 rows=3.00 loops=1)
         Workers Planned: 2
         Workers Launched: 2
         Buffers: shared read=41667
         ->  Partial Aggregate  (cost=67713.03..67713.04 rows=1 width=8) (actual time=164.518..164.519 rows=1.00 loops=3)
               Buffers: shared read=41667
               ->  Parallel Tid Range Scan on bench_tid_range  (cost=0.01..62504.63 rows=2083362 width=0)
                     (actual time=0.095..97.492 rows=1666680.00 loops=3)
                     TID Cond: ((ctid >= '(0,0)'::tid) AND (ctid < '(41667,0)'::tid))
                     Buffers: shared read=41667

Key Improvements:

1. Distributed Cost: The cost estimate dropped from 104k to 68k, indicating the planner correctly anticipates the benefit of parallelism.

2. Shared Load: While the total buffers read (41667) is the same, the work is shared across 3 processes (1 leader + 2 workers), reducing the wall-clock time for the scan itself from ~243ms to ~97ms per worker.

Counter-Intuitive Alternative: Forced Parallel Seq Scan?

You might ask: *"Why didn't Postgres v18 just choose a Parallel Sequential Scan? Wouldn't scanning the whole table with 4 workers be faster than scanning half the table with 1?"*

I tested exactly that on the v18 instance by forcing enable_tidscan = off with 4 workers.

• Execution Time: ~230 ms.

 Finalize Aggregate  (cost=124959.76..124959.77 rows=1 width=8) (actual time=227.917..229.476 rows=1.00 loops=1)
   Buffers: shared hit=15791 read=67543
   ->  Gather  (cost=124959.34..124959.75 rows=4 width=8) (actual time=227.660..229.469 rows=5.00 loops=1)
         Workers Planned: 4
         Workers Launched: 4
         Buffers: shared hit=15791 read=67543
         ->  Partial Aggregate  (cost=123959.34..123959.35 rows=1 width=8) (actual time=224.154..224.155 rows=1.00 loops=5)
               Buffers: shared hit=15791 read=67543
               ->  Parallel Seq Scan on bench_tid_range  (cost=0.00..120834.30 rows=1250017 width=0) 
                     (actual time=0.150..169.772 rows=1000008.00 loops=5)
                     Filter: ((ctid >= '(0,0)'::tid) AND (ctid < '(41667,0)'::tid))
                     Rows Removed by Filter: 999992
                     Buffers: shared hit=15791 read=67543
 Planning:
   Buffers: shared hit=25
 Planning Time: 0.217 ms
 Execution Time: 229.584 ms

It was faster than the serial TID scan (~448 ms)! However, it had to visit all ~83k pages (15,791 hit + 67,543 read) instead of just the ~41k pages we cared about.

The new Parallel TID Range Scan (~150 ms) is still 35% faster than the brute-force/forced Parallel Seq Scan *and* it generates half the I/O load. It is the best of both worlds: fast execution time (Parallelism) and efficient resource usage (Index-like scoping).

Use Case: The Hotel Analogy

To understand why this matters, let's step away from databases and imagine a massive hotel with 1 million rooms. You need to verify all guests whose names start with a letter before 'S'.

The Logical Approach (Offset Problem)

You tell the front desk: *"Give me a list of the first 100 guests with names A-R."* matches 100 people named 'Aaron', 'Abby', etc.

Then you ask: *"Give me the next 100 guests."*

To answer this, the clerk has to skip through the 'A's again just to know where to start counting the 'B's. As you get deeper into the alphabet, the clerk spends more and more time skipping known names just to find the new ones. By the time you are asking for names starting with 'R', they are re-reading 90% of the guest list just to find your batch.

The Physical Approach (TID Scan)

This is the TID Range Scan. You stop caring about the names (the logical value) and organize by Room Number (the physical location).

• Worker 1: "Go check Rooms 1 to 100."
• Worker 2: "Go check Rooms 101 to 200."

Worker 2 doesn't care who is in Room 1. They go straight to Room 101. If the person inside is named "Zach", they skip them. If it's "Alice", they verify them.

• Zero Rescanning: No one re-reads the list of 'A's.
• Parallelism: Because the rooms are physically distinct, you can send 8 workers to different floors simultaneously without them needing to coordinate or skip over each other's results.

This approach transforms an O(N^2) maintainence nightmare into a linear, parallelizable task.

Reproduce this Test

Want to try it yourself on Postgres 19devel? Here is the SQL to set up the test table and run the benchmark.

-- 1. Create the table
DROP TABLE IF EXISTS bench_tid_range;
CREATE TABLE bench_tid_range (id int, payload text);

-- 2. Insert 10M rows to generate ~41k pages
INSERT INTO bench_tid_range 
SELECT x, 'payload_' || x 
FROM generate_series(1, 10000000) x;

-- 3. Vacuum to set visibility map and freeze (important for stable benchmarks)
VACUUM (ANALYZE, FREEZE) bench_tid_range;

-- 4. Enable Parallelism for the Session
SET max_parallel_workers_per_gather = 4; -- Try 2, 4, 8
SET min_parallel_table_scan_size = 0;    -- Force parallel scan even for smaller tables

-- 5. Run the Query
EXPLAIN (ANALYZE, BUFFERS) 
SELECT count(*) 
FROM bench_tid_range 
WHERE ctid >= '(0,0)' AND ctid < '(41667,0)';

Conclusion

This is a welcome "plumbing" improvement. It might not change your daily ad-hoc queries, but for DBAs and developers building custom data maintenance scripts, the ability to parallelize TID based scans is a powerful new tool in the optimization toolkit.

References:

• Commit 0ca3b169
• Discussion Thread

Speed up JOIN Planning - upto 16x Faster!

The hidden cost of knowing too much. That's one way to describe what happens when your data is skewed, Postgres statistics targets are set high, and the planner tries to estimate a join.

For over 20 years, Postgres used a simple O(N^2) loop to compare (equi-join) Most Common Values (MCVs) during join estimation. It worked fine when statistics targets are small (default_statistics_target defaults to 100). But in the modern era - we often see Postgres best-practices recommend cranking that up. Customers are known to be using higher values (1000 and sometimes even higher) to handle complex data distributions + throw a 10 JOIN query to the mix - and this "dumb loop" can easily become a silent performance killer during planning. 

That changes in Postgres 19.

The Problem: It's Quadratic!

When you join two tables, the planner needs to estimate how many rows will match. If both columns have MCV lists (lists of the most frequent values), eqjoinsel() tries to match them up to get a precise estimate.

Historically, it did this by comparing every item in list A with every item in list B.

• If you have 100 MCVs, that's 10,000 comparisons. Fast.
• If you have 10,000 MCVs (max stats target), that's 100,000,000 comparisons. Not so fast.

This meant that simply *planning* a query could take significantly longer than executing it, especially for simple OLTP queries.

The Solution: Hash It Out

The fix is elegant and effective.

Instead of a nested loop, the planner now:

1. Checks if the total number of MCVs is greater than 200 (100 each side).

2. If so, it builds a Hash Table of the MCVs from one (smaller) side.

3. It then probes this hash table with the MCVs from the other side.

The threshold of 200 was chosen because hashing has a startup cost (allocating memory, computing hashes). For smaller lists, the simple loop is actually faster. But once you cross that threshold, the hash table wins.

This transforms the complexity from O(N^2) to O(N), making the estimation step virtually instantaneous even with the largest statistics targets.

Let's Benchmark

I wanted to verify this myself, so I set up a worst-case scenario: two tables with 100,000 rows, but only 10,000 distinct values, and I cranked the statistics target to the maximum (10,000) to force a massive MCV list.

Setup

CREATE TABLE t1 (x int);
CREATE TABLE t2 (x int);

-- Scenario A: 3M rows, Stats 10000
INSERT INTO t1 SELECT x % 10000 FROM generate_series(1, 3000000) x;
INSERT INTO t2 SELECT x % 10000 FROM generate_series(1, 3000000) x;

-- Maximize statistics target
ALTER TABLE t1 ALTER COLUMN x SET STATISTICS 10000;
ALTER TABLE t2 ALTER COLUMN x SET STATISTICS 10000;

ANALYZE t1;
ANALYZE t2;

Results

I ran the following query 10 times on both versions, interleaved to ensure fairness and used median for the results:

EXPLAIN (ANALYZE, TIMING OFF) SELECT count(*) FROM t1 JOIN t2 ON t1.x = t2.x;

Scenario A: High Stats (10k MCVs)

• Rows: 3 Million
• Statistics Target: 10,000
• MCV Count: 10,000

• Before (Postgres 18): ~27.8 ms
• After (Postgres 19): ~1.75 ms
• Speedup: ~16x

Scenario B: Medium Stats (1k MCVs)

• Rows: 300,000
• Statistics Target: 1,000
• MCV Count: 1,000

• Before (Postgres 18): ~0.85 ms
• After (Postgres 19): ~0.60 ms
• Speedup: ~1.4x (40% speed-up)

Scenario C: Default Stats (100 MCVs)

• Rows: 30,000
• Statistics Target: 100 (Default)
• MCV Count: 100

• Before (Postgres 18): ~0.40 ms
• After (Postgres 19): ~0.43 ms
• Speedup: None (Optimization correctly skipped)

Since the total MCV count in Scenario C (100 + 100 = 200) did not exceed the 200 threshold, Postgres 19 correctly chose the simpler loop, avoiding the overhead of building a hash table. This confirms that the patch is smart enough to only kick in when it matters.

The "Quadratic Curve" in Action

To visualize the O(N^2) vs O(N) difference, I ran a benchmark across a wide range of statistics targets (from 10 to 10,000). In each test, the number of rows was set to 300x the statistics target to ensure the MCV list was fully populated and relevant.

Stats Target	PG 18 (ms)	PG 19 (ms)	Speedup
10	0.40	0.38	-
100	0.40	0.45	-
200	0.45	0.43	-
500	0.53	0.52	-
1000	0.85	0.63	1.3x
2000	1.73	0.68	2.5x
5000	7.54	1.14	6.6x
8000	17.97	1.57	11.4x
10000	27.56	1.92	14.3x

As you can see, up to a target of 500, the difference is negligible (and the optimization might not even kick in). But as the target grows, the quadratic cost of the old method explodes, while the new hash-based method scales linearly and remains extremely fast.

This patch is a classic example of modernizing legacy assumptions. The code written 20 years ago assumed MCV lists would be short (in all fairness the default was 100 for a long time). Today's hardware and data requirements have pushed those boundaries, and Postgres is evolving to meet them.

Thanks to Ilia Evdokimov for the patch, David Geier for co-authoring, and Tom Lane for the review.

Note: This feature is currently in the master branch for Postgres 19. As with all development features, it is subject to change before the final release.

References

• Commit: Speed up eqjoinsel() with lots of MCV entries
• Discussion: Mailing List Thread

Teaching Query Planner to See Inside C Functions

The Postgres planner just got a new superpower: X-ray vision for your black-box functions.

For years, Postgres has been able to "inline" simple SQL functions. If you wrote CREATE FUNCTION ... LANGUAGE SQL AS 'SELECT ...', the planner would often rip out the function call and paste the raw SQL directly into the main query. This was magic. It meant that WHERE clauses could be pushed down, indexes could be used, and performance was excellent.

But if you stepped outside the safety of simple SQL functions—say, into the world of C extensions or complex dynamic SQL—you were often stuck with a "black box." The planner saw a function, executed it blindly, and only *then* applied your filters.

That changes in Postgres 19.

The "Black Box" Problem

Imagine you have a function that returns a large dataset, but you only want a few rows:

SELECT * FROM my_complex_function() WHERE id = 42;

Before Postgres 19, if my_complex_function wasn't a simple SQL function, the database would:

1. Run my_complex_function() to completion (generating, say, 1 million rows).

2. Filter the result to find id = 42.

This is painfully inefficient. You wanted an Index Scan, but you got a full execution followed by a filter.

The Solution: SupportRequestInlineInFrom

With this feature, Postgres now allows function authors (specifically in C extensions) to provide a "support function" that tells the planner: *"Hey, if anyone calls me in a FROM clause, here is the raw SQL query tree that represents what I'm doing."*

The planner can then take that query tree and inline it, just like it does for standard SQL functions.

The Impact

This is a game-changer for:

• Extensions: Many extensions provide table-like interfaces (e.g., foreign data wrappers, set-returning functions). They can now expose their internal logic to the planner.
• Dynamic SQL: Functions that generate SQL strings and execute them can now potentially participate in optimization.

Before vs. After

Scenario: A PL/pgSQL function foo_from_bar that executes a dynamic query on text_tbl (containing 1 million rows).

CREATE TABLE text_tbl (f1 text);
INSERT INTO text_tbl SELECT 'row_' || generate_series(1, 1000000);
INSERT INTO text_tbl SELECT 'common_row' FROM generate_series(1, 100);
CREATE INDEX text_tbl_idx ON text_tbl(f1);
ANALYZE text_tbl;

Here is the PL/pgSQL function I used for testing (designed to return 100 rows). Note that it uses dynamic SQL, which is normally opaque to the planner, and then linked it to the C support function test_inline_in_from_support_func using ALTER FUNCTION.

CREATE OR REPLACE FUNCTION foo_from_bar(colname TEXT, tablename TEXT, filter TEXT)
RETURNS SETOF TEXT
LANGUAGE plpgsql
AS $function$
DECLARE
  sql TEXT;
BEGIN
  sql := format('SELECT %I::text FROM %I', colname, tablename);
  IF filter IS NOT NULL THEN
    sql := CONCAT(sql, format(' WHERE %I::text = $1', colname));
  END IF;
  RETURN QUERY EXECUTE sql USING filter;
END;
$function$ STABLE;

ALTER FUNCTION foo_from_bar(TEXT, TEXT, TEXT)
  SUPPORT test_inline_in_from_support_func;

And here is the query I ran:

SELECT * FROM foo_from_bar('f1', 'text_tbl', 'common_row');

Before (PG 18 and earlier)

->  Function Scan on public.foo_from_bar  (cost=0.25..10.25 rows=1000 width=32)
      Output: foo_from_bar
      Function Call: foo_from_bar('f1'::text, 'text_tbl'::text, 'common_row'::text)
      Buffers: shared hit=32 read=4
      Execution Time: 0.353 ms

The function is treated as a black box. The planner estimates 1000 rows (default) and scans the function output.

After (PG 19)

->  Index Only Scan using text_tbl_idx on public.text_tbl  (cost=0.42..8.44 rows=1 width=10)
      Output: text_tbl.f1
      Index Cond: (text_tbl.f1 = 'common_row'::text)
      Buffers: shared hit=1 read=3
      Execution Time: 0.064 ms

The planner "sees through" the function, realizes it's just a query on text_tbl, and uses the index directly. The execution time drops significantly (5.5x faster in this case).

I also ran a pgbench test (1000 iterations) to measure the end-to-end latency improvement:

• Before: 0.107 ms average latency
• After: 0.087 ms average latency
• Improvement: ~19% faster end-to-end (including network overhead)

Bonus: Outer WHERE Pushdown - Orders of magnitude faster!

Next, I ran a test where NULL was passed to the function (returning all 1 million rows) and applied an external WHERE clause:

SELECT * FROM foo_from_bar('f1', 'text_tbl', NULL) WHERE foo_from_bar = 'common_row';

Before (PG 18):

->  Function Scan on public.foo_from_bar  (cost=0.25..12.75 rows=5 width=32)
      Output: foo_from_bar
      Function Call: foo_from_bar('f1'::text, 'text_tbl'::text, NULL::text)
      Filter: (foo_from_bar.foo_from_bar = 'common_row'::text)
      Rows Removed by Filter: 1000000
      Execution Time: 136.267 ms

The database scanned all 1 million rows from the function, then filtered them down to 100.

After (PG 19):

->  Index Only Scan using text_tbl_idx on public.text_tbl  (cost=0.42..8.44 rows=1 width=10)
      Output: text_tbl.f1
      Index Cond: (text_tbl.f1 = 'common_row'::text)
      Execution Time: 0.055 ms

The planner pushed the WHERE clause *inside* the function logic, triggering an index lookup (~2400x speedup).

Similarly, I tested adding a LIMIT 5 clause to the query and saw a ~20x speedup!!

• Before: The database scanned the function output until it got 5 rows (~1.091 ms).
• After: The planner pushed the limit down into the index scan, stopping immediately after finding 5 rows (~0.051 ms).

Testing Methodology

The benchmarks above were conducted using pgbench with the following parameters:

• Iterations: 1000 transactions per client.
• Concurrency: Single client (1 thread).
• Data Volume: The target table text_tbl contained 1,000,100 rows.
• Return Size: The query returned 100 rows per execution.
• Metric: Average latency reported by pgbench (excluding connection overhead).

• Commits tested: c0bc9af (before) and b140c8d (after).

The C support function (test_inline_in_from_support_func) used in this example is defined in src/test/regress/regress.c that is a part of the regression tests. To know more about support functions, see PostgreSQL Documentation on Function Optimization Information.

Discussion & Credits

Coincidentally a Linkedin thread mentioned that Set-Returning Functions (SRF) began their journey in Postgres 7.3 - that was 17 years ago! With this patch, the planner should be able to see through such a function and develop much better execution plans around them! Thanks to Paul A. Jungwirth (for initating the idea with a working patch) and Tom Lane (for in-depth review).

Note: This feature is currently in the master branch for Postgres 19. As with all development features, it is subject to change before the final release.

Settling COUNT(*) vs COUNT(1) debate in Postgres 19

A recent commit to the PostgreSQL master branch brings a nice quality-of-life optimization for a very common SQL pattern - improving performance by up to 64% for SELECT COUNT(h) where h is a NOT NULL column.

If you've ever wondered whether you should use COUNT(*) or COUNT(1), or if you've been dutifully using COUNT(id) on a non-null column, this change is for you.

Note: This feature is currently on the PostgreSQL master branch (committed in November 2025). As with any commit on master, it is subject to change or even rollback before the final release, although this is relatively rare for committed features. If all goes well, this change will be part of the PostgreSQL 19 major version release.

The Optimization

Historically in PostgreSQL, COUNT(*) has been faster than COUNT(1) or COUNT(column).

• COUNT(*): The planner knows it just needs to count rows. It doesn't need to check if any specific column is null.
• COUNT(column): The executor has to fetch the value of the column for each row and check if it is NULL before counting it. This involves "deforming" the tuple (extracting the data), which adds overhead.
• COUNT(1): Often used as a "faster" alternative by users, but treated similarly to COUNT(column) where the "column" is the constant 1.

The new commit, Have the planner replace COUNT(ANY) with COUNT(*), when possible, changes this behavior.

The planner can now detect when you are counting a generic expression (COUNT(ANY)) that:

1. Cannot be NULL: For example, a constant like 1 or a column defined as NOT NULL.

2. Has no extra clauses: No DISTINCT or ORDER BY inside the aggregate.

In these cases, the planner will automatically transform the query to use COUNT(*).

Performance Impact

This optimization eliminates the overhead of tuple deforming and null checking for these cases.

I ran a benchmark on a local PostgreSQL instance to verify the impact, comparing COUNT(h) (not null column), COUNT(1) (constant), and COUNT(*) (rows) on a table with 10 million rows.

Before Patch:

• SELECT count(h) FROM t: ~195 ms
• SELECT count(1) FROM t: ~126 ms
• SELECT count(*) FROM t: ~119 ms
• *Observation:* COUNT(h) is significantly slower (~64% overhead vs count(*)). COUNT(1) is faster than COUNT(h) but still lags behind COUNT(*).

After Patch:

• SELECT count(h) FROM t: ~117 ms
• SELECT count(1) FROM t: ~116 ms
• SELECT count(*) FROM t: ~114 ms
• *Observation:* COUNT(h) sees the major win, improving dramatically to match COUNT(*). COUNT(1) sees a minor speedup, also converging to the optimal COUNT(*) performance.

This means users who prefer COUNT(1) for stylistic reasons, or who count specific non-null columns, will now get the optimal performance of COUNT(*) automatically.

(The patch author, David Rowley, demonstrated similar speedups in the initial discussion).

What about Nullable Columns?

I also tested COUNT(nullable_col) to see if the optimization applies. As expected, it does not.

• Before Patch: COUNT(nullable_col): ~162 ms
• After Patch: COUNT(nullable_col): ~158 ms (No significant change)

Since the planner cannot guarantee the column is free of NULLs, it must still scan and check every value, so the performance remains unchanged. The optimization strictly targets columns where the planner *knows* the value cannot be NULL.

What about DISTINCT or ORDER BY?

I also tested COUNT(DISTINCT h) and COUNT(h ORDER BY h) to see if the optimization applies. As expected, it does not.

• Before Patch: COUNT(DISTINCT h): ~2663 ms (Median of 10 runs)
• After Patch: COUNT(DISTINCT h): ~2708 ms (Median of 10 runs)

Similarly for ORDER BY, the executor still needs to process the values to sort them, so no optimization is possible.

Test Methodology

• Versions:

* Before: Tested on commit dbdc717ac6743074c3a55fc5c380638c91d24afd.

* After: Tested on commit 42473b3b31238b15cc3c030b4416b2ee79508d8c.

• Data: Table t with 10 million rows, populated via INSERT INTO t (h) SELECT 1 FROM generate_Series(1,10000000);.
• Procedure: Each query was executed 10 times in an interleaved pattern (A, B, C, A, B, C...) to ensure fairness and mitigate caching bias. The reported time is the median of those 10 runs.
• System: Intel Core i7-4770 CPU @ 3.40GHz (4 cores/8 threads), 16GB RAM.

Detailed Analysis

To understand *why* the performance improved, we can look at the EXPLAIN plans and table statistics.

• Table Size: The table t (10 million rows) is approximately 346 MB.
• Disk vs. Memory: The benchmark results (median of 10 runs) indicate mostly in-memory / hot-cache performance (~3 GB/s throughput). However, even with some disk I/O involved (as seen in single-run EXPLAIN (ANALYZE, BUFFERS) tests showing ~28k buffer reads), the CPU overhead of deforming the tuple was the dominant differentiator.

The Key Difference:

In the Before scenario, COUNT(h) required the scanner to fetch the column data (width=4), whereas COUNT(*) did not (width=0).

-- Before Patch: COUNT(h)
->  Parallel Seq Scan on t  (... rows=4166657 width=4) ...

In the After scenario, the planner transforms COUNT(h) into COUNT(*), allowing it to skip fetching the column data entirely (width=0).

-- After Patch: COUNT(h)
->  Parallel Seq Scan on t  (... rows=4166657 width=0) ...

This reduction in "width" means the executor doesn't need to extract the integer value from the row, saving significant CPU cycles.

Community Discussion & Trivia

Thanks to David Rowley for working on the patch and Corey Huinker / Matheus Alcantara for reviewing this. The discussions debated on trade-offs between what is fastest and which of these is an "anti-pattern", but it is great to see that now all three are now equivalently optimized.

Future Possibilities:

The infrastructure added (SupportRequestSimplifyAggref) opens the door for further optimizations. For instance, COUNT(NULL) could theoretically be optimized to instantly return 0 without scanning any rows - although that specific change wasn't included in this commit to keep the scope focused.

References

• Commit Message: Have the planner replace COUNT(ANY) with COUNT(*), when possible
• Discussion Thread: Have the planner convert COUNT(1) / COUNT(not_null_col) to COUNT(*)

Pi-hole Part 2: Going Fully Independent with Unbound

After my positive first impressions with Pi-hole, I decided to take the next logical step: eliminating my dependence on external DNS resolvers entirely. While Quad9 served me well, there's something unsettling about routing all my DNS queries through a third party, no matter how trustworthy they appear.

The solution? Unbound - a validating, recursive, caching DNS resolver that can operate completely independently, resolving domain names directly from authoritative sources without relying on upstream DNS providers.

Why Make the Switch?

My motivation goes beyond just privacy paranoia (though that's certainly part of it):

Privacy First: Every DNS query reveals your browsing patterns. Even with Quad9's privacy promises, I prefer keeping that data entirely within my network.

Security Enhancement: Unbound performs DNSSEC validation by default, ensuring the authenticity of DNS responses and protecting against DNS spoofing attacks.

Reduced Latency: While counterintuitive, eliminating the round-trip to external resolvers can actually improve response times for frequently accessed domains through better local caching.

True Independence: No more wondering about logging policies, data retention, or potential government requests to DNS providers.

The Downsides to Consider

Before diving in, it's worth acknowledging the trade-offs:

Increased Complexity: You're now responsible for maintaining and troubleshooting your own DNS infrastructure. When things break, you can't just blame your ISP's DNS servers.

Initial Query Delays: Cold cache scenarios will be slower as Unbound has to walk the entire DNS hierarchy from root servers. First visits to new domains will take noticeably longer.

Resource Usage: While minimal, you're now running an additional service that consumes memory and CPU cycles on your Pi.

Potential Connectivity Issues: If your Pi goes down, your entire network loses DNS resolution. External resolvers provide redundancy that you're giving up.

CDN Sub-optimization: Content delivery networks may not route you to the geographically closest servers, potentially affecting streaming and download performance. Many large DNS providers like Cloudflare and Google use Anycast networks, where the same IP address is announced from multiple geographic locations, automatically routing you to the nearest server. When you run your own recursive resolver, you lose this geographic optimization and might connect to CDN endpoints that are further away.

False Sense of Security: While your DNS queries are now private, it's important to remember that all your actual web traffic (HTTP/HTTPS requests) still flows through your ISP and is visible in their logs. You've privatized the "phone book lookup" but not the actual "conversation" - your ISP can still see which IP addresses you're connecting to, just not the domain names that resolved to them.

Maintenance Overhead: The root hints file should be updated periodically (though it changes infrequently), and you're responsible for keeping your Unbound configuration current and secure.

The Verdict: Benefits Outweigh the Costs

After weighing these trade-offs, the privacy and security benefits still made a compelling case for proceeding. The complexity is manageable for anyone comfortable with basic Linux administration, and the performance impacts are largely theoretical for typical home usage. Most importantly, the peace of mind from knowing exactly where my DNS queries go and how they're handled proved worth the additional overhead.

The Setup Process

Installing and configuring Unbound alongside Pi-hole turned out to be surprisingly straightforward.

Step 1: Install Unbound

sudo apt update
sudo apt install unbound -y

Step 2: Configure Unbound for Security and Performance

I created a custom configuration at /etc/unbound/unbound.conf.d/pi-hole.conf:

sudo nano /etc/unbound/unbound.conf.d/pi-hole.conf

Here's my security-focused configuration:

server:
    # Basic settings
    port: 5335
    do-ip4: yes
    do-ip6: yes
    do-udp: yes
    do-tcp: yes
    
    # Security settings
    trust-anchor-file: "/var/lib/unbound/root.key"
    auto-trust-anchor-file: "/var/lib/unbound/root.key"
    val-clean-additional: yes
    val-permissive-mode: no
    val-log-level: 1
    
    # Privacy settings
    hide-identity: yes
    hide-version: yes
    harden-glue: yes
    harden-dnssec-stripped: yes
    harden-below-nxdomain: yes
    harden-referral-path: yes
    use-caps-for-id: yes
    
    # Performance settings (security-first approach)
    cache-min-ttl: 300
    cache-max-ttl: 86400
    prefetch: yes
    prefetch-key: yes
    
    # Interface settings
    interface: 127.0.0.1
    access-control: 127.0.0.1/32 allow
    access-control: ::1 allow
    
    # Logging
    verbosity: 1
    log-queries: no
    log-replies: no
    
    # Root hints
    root-hints: "/var/lib/unbound/root.hints"

Step 3: Root Hints Management

Root hints are essential files that tell Unbound where to find the DNS root servers - the starting point for all DNS resolution. When installing Unbound via package manager, root hints are included and should be updated automatically through regular system updates.

Note: The root hints file changes infrequently (typically a few times per year when root servers are added, removed, or have IP changes). Rather than manual updates, it's best to keep your system updated through your package manager, which will handle root hints updates appropriately - roughly every 6 months is more than sufficient for most users.

Step 4: Initialize DNSSEC Root Key (Usually Optional)

Modern Unbound packages typically handle DNSSEC root key initialization automatically. However, if you encounter DNSSEC validation errors or want to ensure the key is properly configured, you can initialize it manually:

sudo unbound-anchor -a "/var/lib/unbound/root.key"
sudo chown unbound:unbound /var/lib/unbound/root.key

Note: If this step fails or the file already exists, it's likely already configured correctly by the package installation.

Step 5: Start and Enable Unbound

sudo systemctl enable unbound
sudo systemctl start unbound

Step 6: Test Unbound

Before integrating with Pi-hole, I verified Unbound was working with both valid and invalid domain lookups:

Testing with an existing domain:

robins@pi4:~ $ dig @127.0.0.1 -p 5335 google.com | egrep -w 'status|^google'
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 58713
google.com.             269     IN      A       142.250.70.206

Testing with a non-existent domain:

dig @127.0.0.1 -p 5335 asdfasdsfasfdfsafasd | egrep -w 'status|flags'

robins@pi4:~ $ dig @127.0.0.1 -p 5335 asdfasdsfasfdfsafasd | egrep -w 'status|flags'
;; ->>HEADER<<- opcode: QUERY, status: NXDOMAIN, id: 19691
;; flags: qr rd ra ad; QUERY: 1, ANSWER: 0, AUTHORITY: 1, ADDITIONAL: 1
; EDNS: version: 0, flags:; udp: 1232

Both tests confirm Unbound is working correctly. The existing domain test shows successful resolution with NOERROR status, while the non-existent domain test returns NXDOMAIN (Non-eXistent DOMAIN), which is the proper response when a domain doesn't exist. Note the ad flag in the non-existent domain response, indicating that even failed lookups are DNSSEC-validated - Unbound cryptographically verified that this domain truly doesn't exist rather than just accepting an unvalidated negative response.

Step 7: Configure Pi-hole to Use Unbound

In the Pi-hole admin interface:

1. Navigate to Settings → DNS
2. Uncheck all upstream DNS servers
3. Add 127.0.0.1#5335 as a custom DNS server
4. Enable DNSSEC validation
5. Save settings

The Results: Worth the Effort

After 24 hours of operation with the new setup, the benefits are clear:

Complete Privacy: All DNS queries now resolve independently without touching external servers (except for the initial root server queries, which reveal no specific domain information).

DNSSEC Validation: Every response is cryptographically verified, providing protection against DNS manipulation.

Improved Cache Efficiency: Unbound's more sophisticated caching algorithm seems to provide better hit rates for our household's browsing patterns.

Performance Impact: Negligible. The Pi 4 handles both Pi-hole and Unbound without breaking a sweat:

robins@pi4:~ $ uptime
 21:22:25 up 2 days,  9:47,  3 users,  load average: 0.00, 0.00, 0.00

Security Considerations

My configuration prioritizes security over raw performance:

• Conservative TTL settings: Shorter cache times mean more frequent validation
• Strict DNSSEC validation: No permissive mode that might accept invalid responses
• Minimal logging: No query logging to preserve privacy even locally
• Restricted access: Only localhost can query Unbound directly

Final Thoughts

Setting up Unbound with Pi-hole represents the logical conclusion of taking control over your DNS infrastructure. While the privacy and security benefits are the primary motivators, the technical satisfaction of running a completely self-contained DNS resolution system is considerable.

The setup process is more involved than simply pointing Pi-hole at an external resolver, but the configuration is straightforward and well-documented. For anyone concerned about DNS privacy or wanting to reduce external dependencies, this combination provides an excellent solution.

The Pi 4 continues to prove itself as the perfect platform for this type of network infrastructure project - handling both Pi-hole and Unbound with resources to spare. As I mentioned in part 1, even a Pi 2 would likely suffice for most households, making this an accessible upgrade for anyone looking to enhance their home network's privacy and security posture.

Pi-hole First Impressions: Home DNS Caching Done Right

(This is first of a 2 part series, where I explore pi-hole as a DNS caching solution-Here's Part 2)

I've had a Raspberry Pi 4 sitting idle for months, and finally decided to put it to good use by setting up Pi-hole for network-wide DNS caching and ad blocking. After less than a day of operation, I'm genuinely impressed with both the installation process and the results.

The Installation: Surprisingly Smooth

Setting up Pi-hole turned out to be one of the smoothest software installations I've experienced in recent memory. The entire process is handled by a single bash script:

curl -sSL https://install.pi-hole.net | bash

That's it! What impressed me most wasn't just the simplicity, but the thoroughness of the pre-installation checks. Before the script even attempted to install anything, it ran through a comprehensive validation process:

• Network connectivity tests
• DNS resolution verification
• Package manager status checks
• System requirements validation
• Port availability confirmation

This rigorous validation gave me real confidence that the installation would succeed. Too often, installation scripts fail halfway through, leaving you with a partially configured mess. Pi-hole's approach of "check everything first, then install" meant that once the green lights were all showing, the installation proceeded flawlessly.

Interestingly, what took the most time during the entire setup wasn't Pi-hole itself, but my initial attempt to use Docker. I'd assumed Docker would be the quick and easy route (as it usually is), but the Pi had some leftover Docker installations that needed cleaning up first - removing old docker, docker.io, and docker-compose packages was a messy affair. Then the Pi seemed to have issues with the Docker installation process itself. In the end, I abandoned the Docker approach and went with the native installation, which ironically turned out to be much faster and cleaner.

One lesser-known tip that made the physical setup even smoother: if you're running a Google mesh network, the routers with both WAN and LAN ports can easily be used as wifi-to-LAN bridges. Rather than connecting the Pi 4 via WiFi, I simply plugged it into the LAN port of one of the mesh routers. This eliminated the need for additional network switches and reduced latency by a few milliseconds by avoiding the WiFi overhead entirely.

Another major convenience: you don't need to update DNS settings on each individual device. A single configuration change at the router level does the job for the entire household. In Google mesh, I simply changed the DNS setting from "Automatic" to "Custom" and set the Pi 4's IP as the primary DNS server. This immediately moved all devices on the network to use Pi-hole without touching any individual device settings. The only exceptions are devices that explicitly force their own DNS (looking at you, certain smart TVs and streaming devices), but those are relatively rare.

The Results: Better Than Expected

After 23 hours of operation protecting our home network, the statistics are eye-opening:

DNS Query Volume: 100,000 requests in 23 hours

• That's roughly 4,350 DNS queries per hour from our household
• Averages out to about 72 queries per minute
• Shows just how chatty modern devices really are

Cache Performance: 70% cache hit rate

• Pi-hole is successfully caching DNS responses locally
• Reduces latency for frequently accessed domains
• Decreases load on upstream DNS servers
• Currently using Quad9 (9.9.9.9) as the upstream resolver - the only non-US DNS service I still trust in this day and age

Blocking Effectiveness: 8.5% of requests blocked

• More than 1 in 12 DNS requests were for ad/tracking domains
• That's 8,500 blocked requests in just 23 hours
• Represents a significant reduction in unwanted network traffic

System Performance: Negligible Impact

One concern with running Pi-hole on a Pi 4 was whether it would impact system performance. For context, here are the system specs:

robins@pi4:~ $ uname -a
Linux pi4 6.12.34+rpt-rpi-v8 #1 SMP PREEMPT Debian 1:6.12.34-1+rpt1~bookworm
  (2025-06-26) aarch64 GNU/Linux

robins@pi4:~ $ free -h
               total        used        free      shared  buff/cache   available
Mem:           3.7Gi       463Mi       1.7Gi        34Mi       1.6Gi       3.3Gi
Swap:          511Mi          0B       511Mi

The performance numbers are reassuring:

CPU Usage: Essentially zero load:

robins@pi4:~ $ uptime
 10:39:33 up 23:04,  3 users,  load average: 0.00, 0.00, 0.00

• Zero computational overhead even after 23+ hours of operation
• Pi 4 is more than capable of handling the workload
• Plenty of headroom for additional services

Memory Usage: Minimal footprint

• Pi-hole runs efficiently even on modest hardware
• No noticeable impact on system responsiveness

Honestly, the Pi 4 seems like overkill for this task. Given the minimal resource requirements, a Pi 2 (with the older Cortex-A7) would likely handle Pi-hole just fine, leaving the Pi 4 (with the more capable Cortex-A72) free for more power-hungry side projects. The beauty of Pi-hole is that it's so lightweight, you can run it on practically any Raspberry Pi model and still have resources to spare.

The Web Interface: Polished and Responsive

Pi-hole's web dashboard deserves special mention. It's:

• Fast and Responsive: Page loads are snappy, even on the Pi 4
• Live Updates: The dashboard shows real-time query statistics
• Well-Designed: Clean interface that makes data easy to understand
• Comprehensive: Detailed logs, statistics, and configuration options

The live dashboard is particularly satisfying to watch - seeing blocked queries in real-time gives you a visceral sense of how much unwanted traffic Pi-hole is filtering out.

Real-World Benefits

Beyond the statistics, the practical benefits are noticeable:

• Faster Browsing: Pages load quicker without ad network delays
• Cleaner Experience: Websites feel less cluttered
• Privacy Improvement: Reduced tracking across devices
• Bandwidth Savings: Less unwanted traffic on the network

One particular benefit I'm looking forward to testing: my network occasionally suffers from stuttering ping latencies. Instead of the usual 10ms responses, I'll sometimes see 100ms or even 1-second ping times that persist for minutes. During these episodes, streaming mostly works and browsing barely functions, but what really hurts is DNS resolution - especially for websites requiring multiple round-trips. With Pi-hole's 70% cache hit rate, those problematic periods should be noticeably less laggy since most DNS queries won't need to traverse the struggling network connection. More testing needed, but the potential is promising.

The ad filtering aspect wasn't my primary motivation, but it doesn't hurt either. For websites I frequent that provide genuinely good content - like Phoronix with their constant stream of quality tech updates - I've set up pass-throughs so they can rightfully earn the advertising revenue they deserve for their work. For projects like Pi-hole itself, I prefer the direct contribution route via donations. It's about supporting the creators and maintainers who provide value, whether through allowing ads or direct financial support.

Worth Supporting

The Pi-hole project has created something genuinely useful that "just works" out of the box. The quality of both the software and the installation experience makes this a project worth supporting financially. I'll definitely be making a donation to help ensure continued development.

Final Thoughts

Pi-hole represents the best kind of open-source software: it solves a real problem elegantly, installs without drama, and delivers measurable benefits immediately. If you have a spare Raspberry Pi lying around, setting up Pi-hole is an excellent way to put it to work improving your entire network's performance and privacy.

The fact that it blocked 8,500 unwanted requests in less than a day while using virtually no system resources makes it a clear win. Sometimes the best technology is the kind you set up once and then forget about - until you look at the statistics and realize how much work it's quietly doing in the background.

Taming ReorderBufferWrite - Boost Logical Decoding in Postgres

Performance bottlenecks in Postgres logical replication or Change Data Capture (CDC) stream can be subtle, but one specific wait event, ReorderBufferWrite, often points directly at how your application interacts with the database during these processes. Let's unpack this wait-event and see how your application's workload patterns can trigger it.

Core Concepts: Logical Decoding and WAL

To understand the problem, we first need some context:

• Write-Ahead Log (WAL) is PostgreSQL's transaction log and every change is written here first (to ensure durability and enable recovery/replication)

• Logical Decoding is a powerful feature that reads this WAL. Instead of just replaying physical changes (like physical replication), it translates the WAL records into a logical, easy-to-understand stream of changes, carefully taking care of rows and transactions, which is then consumed by logical replication subscribers or CDC systems.

The Reorder Buffer: Ensuring Transactional Order

Logical decoding needs to present changes in the exact order transactions were committed. However, changes from different concurrent transactions are interleaved within the WAL. To solve this, PostgreSQL uses an in-memory area called the Reorder Buffer. Its job is to collect decoded changes belonging to transactions that are still in progress. Only when a transaction commits are its changes released from the buffer in the correct sequence.

Analogy: The Assembly Line QC Station

For an analogy, imagine an assembly line where components (WAL entries) arrive continuously where workers assemble different products (transactions). Some products are quick builds, others complex multi-step assemblies. Completed products move to a Quality Control (QC) holding station (Reorder Buffer). The QC inspector (logical decoding process) ships out batches of products, but only when all products belonging to a specific batch (committed transaction) have passed inspection. The batches must be shipped in the strict order they were completed.

The Bottleneck: ReorderBufferWrite

The Reorder Buffer uses memory allocated by the logical_decoding_work_mem setting (often defaulting to 64MB). What happens if this QC holding station (Reorder Buffer) gets completely filled up with products waiting for their batch-mates (changes from active transactions), especially for those complex, slow builds (long-running transactions)?

The system can't just discard these waiting items and the best it can do is to move them temporarily to an overflow warehouse (spilling to disk). This action – writing the buffer's contents to slower disk storage – is precisely the ReorderBufferWrite wait event.

Waiting for disk I/O is significantly slower than operating in memory. Frequent ReorderBufferWrite events directly translate to increased replication lag and reduced throughput for the logical change stream. Retrieving items from the overflow warehouse (disk) drastically slows down the QC inspector's (logical decoding) shipping rate.

Workloads That Cause Spills to Disk

ReorderBufferWrite waits are primarily triggered by workload patterns that overwhelm the in-memory Reorder Buffer. Key culprits include:

1. Very Large Transactions: A single transaction modifying hundreds of thousands or millions of rows generates a massive volume of changes that must sit in the Reorder Buffer until the final commit. This can rapidly exhaust the available memory. Think of assembling a huge, complex product requiring many components – it occupies a lot of space at the QC station while being built.

2. Long-Running Transactions: A transaction might not change that much data, but if it stays open for a long time (perhaps due to complex calculations, waiting for external input, or slow queries within it), its changes linger in the Reorder Buffer. Meanwhile, other transactions complete, adding their changes. The buffer fills up with changes from many transactions, bottlenecked by the one(s) taking a long time to commit. This is like one single (inactive / idling) product assembly holding up the shipment of multiple completed batches at the QC station.

3. High Volume of Concurrent Changes: Even with reasonably sized and timed transactions, a very high rate of change across many concurrent sessions can collectively generate data faster than the logical decoding process can handle within the allocated Reorder Buffer memory, leading to spills. Imagine hundreds of small, quick products arriving at the QC station so fast it still gets overwhelmed.

• Sudden workload spike: A related reasoning is when there is a sudden surge of application workload (writes) and they cause a rate of change higher than the logical decoding process can generally handle. In terms of workload pattern, to an experienced DBA, this may appear as:

• connection spike => followed by replica-lag => and thereafter a slow recovery of that replica lag.

4. Insufficient logical_decoding_work_mem: If this setting is too low for your workload's typical peak demands, spills will occur more readily, even for moderately busy workloads.

What need be done?

As an application developer, you have significant control over the transaction patterns if ReorderBufferWrite waits are causing issues:

1. Audit & Split Large Transactions: Review bulk DML operations performed within a single transaction and refactor these into smaller batches. For e.g. Instead of one transaction for say 100k rows, use 10 transactions of 10k rows each. This drastically reduces the peak Reorder Buffer usage.

2. Minimize Transaction Duration: Identify code paths where transactions are held open unnecessarily long. This could be from end-user applications waiting for user-input, or doing non-database work, or calling slow external services inside a database transaction that modifies replicated tables. Try to keep database transactions short and focused. Perform reads, calculations, or external calls before starting the write transaction, or after it commits. 

3. Optimize Queries Within Transactions: Slow SELECT, INSERT, UPDATE, or DELETE queries will naturally extend transaction duration. Use query analysis tools (EXPLAIN ANALYZE) to find and optimize slow queries involved in your transactions.

4. Isolate Read-Only Work: If long-running processes only read data, ensure they aren't unnecessarily starting read-write transactions on the primary where logical decoding is happening. Similarly use read-only transactions or more permissive isolation levels where possible.

5. Collaborate on Configuration: If application-side optimizations aren't sufficient, discuss the observed workload with your DBA or platform team. It would help to provide them with information about your transaction patterns. This could help them review whether an increase in logical_decoding_work_mem is warranted and safe within the server's overall resource constraints. Tuning memory should usually follow workload optimization.

Conclusion

ReorderBufferWrite isn't just an obscure PostgreSQL internal wait-event. It's a performance indicator directly linked to application workload patterns during logical decoding. By designing your application to use shorter, smaller, and more efficient transactions, especially on tables involved in replication, you can minimize these disk spill events and ensure a performant database.