Brendan Gregg

SEPTEMBER 25, 2021

A Cassandra database cluster had switched to Ubuntu and noticed write latency increased by over 30%. CLI tools The Cassandra systems were EC2 virtual machine (Xen) instances. Note that Ubuntu also has a frame to show entry into vDSO (virtual dynamic shared object). Microbenchmark os::javaTimeMillis() on both systems.

Speed 126

Speed Java AWS Virtualization 126

HammerDB

MAY 5, 2019

Fortunately the HammerDB TPC-C/OLTP workload intentionally has a great deal of contention between threads and is therefore ideal for testing spin-locks. ” and “as the PAUSE latency has been increased significantly, workloads that are sensitive to PAUSE latency will suffer some performance loss.” linux-glibc2.12-x86_64/data

Testing 48

Testing Tuning Latency Storage 48

HammerDB

OCTOBER 18, 2019

cpupower frequency-info analyzing CPU 0: driver: intel_pstate CPUs which run at the same hardware frequency: 0 CPUs which need to have their frequency coordinated by software: 0 maximum transition latency: Cannot determine or is not supported. bin/pgbench -c 1 -S -T 60 pgbench starting vacuum.end. hardware limits: 1000 MHz - 4.00

Benchmarking 63

Benchmarking Performance Hardware C++ 63

Brendan Gregg

OCTOBER 8, 2018

more capable, and built from the ground up for the modern era of the eBPF virtual machine. Screenshot: tracing read latency for PID 181: # bpftrace -e 'kprobe:vfs_read /pid == 30153/ { @start[tid] = nsecs; } kretprobe:vfs_read /@start[tid]/ { @ns = hist(nsecs - @start[tid]); delete(@start[tid]); }'. Attaching 2 probes. ^C

C++ 110

C++ Virtualization Programming Latency 110

Brendan Gregg

SEPTEMBER 25, 2021

A Cassandra database cluster had switched to Ubuntu and noticed write latency increased by over 30%. CLI tools The Cassandra systems were EC2 virtual machine (Xen) instances. Note that Ubuntu also has a frame to show entry into vDSO (virtual dynamic shared object). Microbenchmark os::javaTimeMillis() on both systems.

Speed 40

Speed Java AWS Virtualization 40

HammerDB

OCTOBER 12, 2018

This post complements the previous best practice guides this time with the focus on MySQL and MariaDB and achieving top levels of performance with the HammerDB MySQL TPC-C test. SELECT DISTINCT c from sbtest where id between ? order by c. SELECT c from sbtest where id=? Copyright (C) 2003-2018 Steve Shaw.

Best Practices 40

Best Practices Scalability Performance C++ 40

Brendan Gregg

OCTOBER 7, 2018

more capable, and built from the ground up for the modern era of the eBPF virtual machine. Screenshot: tracing read latency for PID 181: # bpftrace -e 'kprobe:vfs_read /pid == 30153/ { @start[tid] = nsecs; } kretprobe:vfs_read /@start[tid]/ { @ns = hist(nsecs - @start[tid]); delete(@start[tid]); }'. Attaching 2 probes. ^C

C++ 40

C++ Virtualization Programming Latency 40