Explore tweets tagged as #pgbench
PostgreSQL benchmarks on @akashnet providers You can use pgbench, PostgreSQL’s built-in benchmarking tool, to measure real-world latency: • pgbench -i -s 1 -h {PROVIDER} -p 31736 -U admin -d mydb → initializes the database and populates it with test data based on the scale
1
2
3
Rintaro Ikeda of NTT Data is on stage here at #PGConfdev giving a lightning talk titled, "My First Patch – adding new option to pgbench" & everyone who has run performance benchmarks in #PostgreSQL is paying attention
0
3
16
PlanetScale’s $5 Postgres vs my Hetzner CPX11 (same pgbench mix, eu-central, direct vs PgBouncer). Latency hurts, pooling helps, charts included. (See first reply!)
2
0
1
At #PGConfNYC @PGConfNYC right now, @melanieplageman is on stage giving a talk on “Introduction to benchmarking with pgbench” and as always Melanie is good at EXPLAINing all things #PostgreSQL
2
6
45
New Intel 14900K running in my benchmark lab. First of my personal hardware to hit 1M mini SELECT/sec: $ pgbench -S -T 30 -c 32 -j 32 pgbench tps = 1053277.342735 Repeating that eventually caused a reboot crash. Probably need to dampen the freq/power limits to stabilize it.
1
1
9
🕵️♂️“kubectl debug” is a Swiss-army knife for investigations. My favorite trick: run pgbench inside the same Pod as Postgres. It eliminates all network noise from the test. Other surprisingly handy moves: kubectl debug -c gdb … → drop a container with gdb, strace, tcpdump —
1
4
23
AWS,Azure,GCP,SupabaseのマネージドPostgreSQLを比較した記事。同スペックのインスタンスでのpgbenchの結果はRDSが圧勝。ノイジーネイバー問題とかあるのでタイミングに左右される気がするけど結構な差があるように見える。 https://t.co/ZQz7oeJ96G
2
4
15
PostgreSQL 17 dev: Allow pgbench to exit immediately when any client is aborted. https://t.co/jjlV7pJs7I
0
1
12
2 million TPS Postgres 16, SELECT-only pgbench (w/o prepared statements), 4th-gen AMD EPYC with 360 vCPUs in GCP https://t.co/vtu0WhlibN
2
5
32
There are things where LLM hallucination is actually useful, suggesting ideas or highlighting issues e.g. in old Postgres versions (here -- 14), in pgbench, the option "-d" didn't mean "database", but rather "debug" I confused myself in the very same way so many times for years
0
0
6
PostgreSQL 16のpsql/pgbenchでロードバランスを試してみる https://t.co/GYtaF0L2sP
0
3
27
Desktop bare metal PostgreSQL 16.2 clearing 10GB/s reads here with a PCIe 5.0 @CrucialMemory T705. i9-14900K, Ubuntu 23.10, 32 clients @~325MB/s each. select-pages has hard to cache long index range scans, it's one of my pgbench-tools benchmark workloads to stress SSD.
0
0
5
Ready to take your Postgres performance to the next level? EDB’s Mark Wong dives deep into the world of complex #Postgres benchmarking using TPC-C and #Pgbench. Unlock the full potential of your Postgres database with our comprehensive guide. https://t.co/geAaeR1yLP
0
0
5
How to run pgbench through a DataSunrise proxy for PostgreSQL databases. https://t.co/1gVsuRC7Po
#PostgreSQL #proxy #databaseprotection
0
0
0
I love ASH maths with a two-layer architecture. If you PgBench from 5 clients single row inserts, you see 5 active sessions on the SQL layer in "Query Processing" and 5 active threads "OnCpu_Active" in the storage layer. It also matches what you see in pg_stat_activity.
0
0
9
First time I see 3 million TPS on a single machine with @PostgreSQL!! A SELECT-only pgbench experiment conducted by @postgres_ai
https://t.co/TGndzPMKHS (don't rush to jump into conclusion that pgss has big overhead – this is an edge case: fast SELECTs, prepared statements, a
2
7
28
PostgreSQL 19 dev: pgbench: Add --continue-on-error option. Author: @rikeda361
https://t.co/ZuqSpVTWAm
0
0
18
