Unconventional PostgreSQL Optimizations(hakibenita.com) |
Unconventional PostgreSQL Optimizations(hakibenita.com) |
This is part of a broader choice: write amplification. You'd want to, of course, have the most precise index possible - but no matter how you cut it, you are incurring extra I/O for writes - one for the tuple, one per index. How you index things is heavily influenced by the mix of reads and writes, and this is why we have data warehouses/read replicas in the first place: it allows us to avoid write amplification in the write path, while having fast filtered reads (that are slightly delayed).
If you're dealing with <ridiculous number of users>, there is a good chance that you don't want to be putting BI/OLAP indices on your OLTP database. You probably don't have enough users to worry about this - but - if you ever find that your writes are becoming an issue this is something to consider.
However, that is usually the least of your concerns with write amplification. If you don't batch your writes, you can easily get 100x write amplification. For any primary key or any other index not strongly correlated with your INSERTs, you can get perhaps another 100x write amplification even if you batch you writes.
But it would add complexity to detect out-of-sync indexes and tables.
> I'm not a big fan of using the constraint names in SQL, so to overcome both limitations I'd use MERGE instead:
``` db=# MERGE INTO urls t USING (VALUES (1000004, 'https://hakibenita.com')) AS s(id, url) ON t.url = s.url WHEN MATCHED THEN UPDATE SET id = s.id WHEN NOT MATCHED THEN INSERT (id, url) VALUES (s.id, s.url); MERGE 1 ```
I use `insert ... on conflict do update ...` all the time to handle upserts, but it seems like merge may be more powerful and able to work in more scenarios. I hadn't heard of it before.
https://pganalyze.com/blog/5mins-postgres-15-merge-vs-insert...
This is somewhat a personal preference, but I would just use `INSERT ... ON CONFLICT` and design my data model around it as much as I can. If I absolutely need the more general features of `MERGE` and _can't_ design an alternative using `INSERT ... ON CONLFICT` then I would take a bit of extra time to ensure I handle `MERGE` edge cases (failures) gracefully.
> If you want the generality of MERGE, you have to accept the fact that you might get unique constraint violations, when there are concurrent inserts, versus with INSERT ON CONFLICT, the way it's designed with its speculative insertions, guarantees that you either get an INSERT or an UPDATE and that is true even if there are concurrent inserts. You might want to choose INSERT ON CONFLICT if you need the guarantee.
Basically, `MERGE` is susceptible to a concurrent process also writing `INSERT` where that `INSERT` and `MERGE` are unaware of one another, causing a duplicate value to be used.
Besides portability, there is IMHO nothing against INSERT ... ON CONFLICT if it does what you need.
Since I know conceptually how RDBMSes work, I can ask vey specifically what I want. Also asking for feedback on schemas/queries really helped me. I use a lot more of PGs features now!
How well do they work for UUIDv7? You’d probably have to tune (increase?) pages_per_range, but even though each index entry is 16 bytes you have to consider the btree index on the same is also similarly affected (or worse).
It's interesting how both virtual columns and hash indexes work, but feel like they're bolted on, vs being made part of the whole ecosystem so that they work seamlessly.
Hash indices have long been crippled; they shipped almost unusable but every few years get a good QoL update. I think automatic unique constraints are the big thing left there.
> Starting at version 14, PostgreSQL supports generated columns - these are columns that are automatically populated with an expression when we insert the row. Sounds exactly like what we need but there is a caveat - the result of the expression is materialized - this means additional storage, which is what we were trying to save in the first place!
Is it also possible to create index (maybe partial index) on expressions?
Postgres makes DELETEs single threaded, this includes the selection part of the DELETE. By running a completely separate SELECT first Postgres would multithread the SELECT and populate the cache fast. Then the single thread DELETE can operate on in-memory data and not endlessly block loading data from disk.
PG's lack of plan caching strikes again, this sort of thing is not a concern in other DB's that reuse query plans.
The article suggests using a check constraint to get around that - are you saying that does actually check the underlying value when USING HASH is set? If so, I think the docs need updating.
Of course the hash index also outperforms a unique (btree) index on top of separately calculating the hash, in addition to the storage overhead, row bloat, lack of guarantees regarding the hash unless you expose it to Postgres as a user-defined function AND add a check constraint.
Is the syntax highlighting built into pgsql now or is that some other wrapper that provides that? (it looks really nice).
My only gripe with it is its insistence on adding a space after a line break when the query is too long, making copy/paste a pain for long queries.
It sometimes really stinks on some queries since the generic plan can't "see" the parameter values anymore. E.g. if you have an index on (customer_id, item_id) and run a query where `customer_id = $1 AND item_id = ANY($2)` ($2 is an array parameter), the generic query plan doesn't know how many elements are in the array and can decide to do an elaborate plan like a bitmap index scan instead of a nested loop join. I've seen the generic plan flip-flop in a situation like this and have a >100x load difference.
The plan cache is also per-connection, so you still have to plan a query multiple times. This is another reason why consolidating connections in PG is important.
0: https://www.postgresql.org/docs/current/runtime-config-query...
MSSQL server also does parameter sniffing now days and can have multiple plans based on the parameters values it also has a hint to guide or disable sniffing because many times a generic plan is actually better, again something else PG doesn't have, HINTS [2].
PG being process based per connection instead of thread based makes it much more difficult to share plans between connections and it also has no plan serialization ability. Where MSSQL can save plans to xml and they can be loaded on other servers and "frozen" to use that plan if desired, they can also be loaded into plan inspection tools that way as well [3].
1. https://learn.microsoft.com/en-us/sql/relational-databases/n...
2. https://learn.microsoft.com/en-us/sql/t-sql/queries/hints-tr...
3. https://learn.microsoft.com/en-us/sql/t-sql/queries/hints-tr...
If you're experiencing things that smell like TOCTOU, first you need to be sure you don't have oddball many-to-one issues going on (i.e., a cardinality violation error), and then you're going to have to increase your transaction isolation level to eliminate non-repeatable reads and phantom reads.
Like, the alternative to a MERGE is writing a few UPDATE statements followed by an INSERT and wrapping the entire batch in a transaction. And you should likely still wrap the whole thing in a transaction. If it breaks, you just replay the whole thing. Re-run the whole job.
At read committed (default) isolation level, INSERT ... ON CONFLICT handles concurrent, conflicting inserts just fine, while MERGE ... WHEN NOT MATCHED (e.g.) does not. This is surprising behavior from the syntax alone, one would assume the two statements, when written with the same intent, would have the same behavior. Of course, this difference is documented, see the last paragraph of the Notes section on MERGE: https://www.postgresql.org/docs/18/sql-merge.html#id-1.9.3.1...
I don't know this for sure, but I believe that the effect of raising the transaction isolation level will just be to turn the constraint violation into a serialization error. That's not any easier to handle gracefully.
The CLUSTER command in PG just moves rows around in the heap so they match the still separate index order which can help a little bit with range operations because rows are close on disk, but otherwise doesn't do much.
So they are completely separate things that just happen to use the same term.
Not sure I follow most tables are accessed primarily in one way (primary key) while maybe sometimes in others for analysis. Having the PK written twice because it's almost always indexed is normally a waste and good candidate for a clustered index. So much so that many DB's like SQLite and MySql always do clustered indexes on primary key because their storage engine is built such that tables are a b-tree anyway vs PG that has separate b-tree indexes and heap tables. MSSQL and Oracle give you a choice whether the table is a index structure or a heap.
If you have very specific use case tables they can typically have a clustered index and no secondary indexes, you can still scan them for ad-hoc analysis but you get better insert performance and space usage because you aren't double writing to the heap and a PK index like you would in PG.
As far as batch writes that is a separate issue and has to due with whether that even makes sense for durability, if you need to commit a single random row due to something occurring you can't batch that up and maintain consistency, if your bulk loading data sure and is common practice to do commit batches there, clustered indexes could still be a 100 vs 200x write amplification if you have to insert both an index row and heap row vs just a single clustered index row.
OTOH, with a heap-less (aka. clustered, aka. index organized) table, you would only have to update the indexes containing the columns that are actually being updated. You don't need to touch any other index. Furthermore, only if you are updating a key column would you physically "move" the entry into a different part of the B-tree. If you update an included column (PK columns are automatically "included" in all secondary indexes, even if not explicitly mentioned in the index definition), you can do that in-place, without moving the entry.
Here is how this works in SQL Server - consider the following example:
CREATE TABLE T (
ID int,
NAME nvarchar(255) NOT NULL,
AMOUNT int NOT NULL,
CONSTRAINT T_PK PRIMARY KEY (ID)
);
GO
CREATE INDEX T_I1 ON T (NAME);
GO
CREATE INDEX T_I2 ON T (AMOUNT);
UPDATE T SET AMOUNT = 42 WHERE ID = 100;
However, seeking into a secondary index, and then reading a column not included in that index incurs an additional index seek (into the clustered index), which may be somewhat slower than what would happen in a heap-based table.
So there are pros and cons, as usual...
Remember some databases always use clustered index internally (SQLite, MySql) such that even if you have no primary key they will create a hidden one instead for use with the index.
https://www.sqlite.org/rowidtable.html
It is nice to have the choice which way to go and would be nice if PG implemented this. It can have significant space savings on narrow table with one primary index and performance advantages.
One possible reason is that the planner configuration can be different per connection, so the plans might not transfer
I believe the plan data structure PG is intimately tied to process space memory addresses since it was never thought to share between them and can even contain executable code that was generated.
This makes it difficult to share between processes without a heavy redesign but would be a good change IMO.
> One possible reason is that the planner configuration can be different per connection, so the plans might not transfer
That's part of it, another big part is that the transactional DDL makes it more complicated, as different sessions might require different plans.
Would be interesting for indexes say put them on ram drive and rebuild them on restart if they aren't there just fallback to table scans.
MSSQL has memory optimized tables that do this sort of thing: https://learn.microsoft.com/en-us/sql/relational-databases/i...
I don’t know if or how Postgres records the transaction number in the index to be able to notice if it’s out of date. If it does, I don’t know of any solution to “catch up” the index besides recreating it, which would be ok if that’s the only issue but from my experience with out-of-date indexes (libc or icu updates, where Postgres doesn’t know if anything IS broken and just reports that it could be), there’s no guarantee you’d even notice and your app could be running completely broken until you rebuild.
That is not my understanding:
https://www.postgresql.org/docs/current/app-pgresetwal.html
>After running this command on a data directory with corrupted WAL or a corrupted control file, it should be possible to start the server, but bear in mind that the database might contain inconsistent data due to partially-committed transactions. You should immediately dump your data, run initdb, and restore. After restore, check for inconsistencies and repair as needed.