Articles in Performance
NFS I/O Stats with Logging
The nfsiostat is an excellent tool for analyzing NFS performance. The only major problem with this utility is that it’s designed to be used interactively. It does not have logging or timestamping capabilities. I needed …
Understanding Memory Utilization in Linux
This is a depressing – and all-too-common – scenario: a user runs the free command and opens a support case because he sees zero “free” memory on the server.
Analyzing atop Logs with atopsar
I have discussed atop previously but concentrated primarily on how to run it and how to collect data. Now I’d like to spend some time talking about ways to analyze the data collected with atop.
Analyzing Network Performance
Much of network performance analysis will be comparative in nature. Thus, seeing the output of multiple commands side by side can be quite useful. Bash has a useful little utility called pr and we’ll make …
Measure DNS Server Performance
This is not an entirely proper way to benchmark a DNS server, but, in a pinch, it should give you some idea of its responsiveness and stability.
Atop Script with Scheduling and Logging
When something is going down on a server, the first thing most sysadmins will run is the venerable top utility. This happens automatically: if you suspect the server is being sluggish, your fingers just type top without you …
Reduce SNMPd Logging Level
By default, (on RHEL/CentOS, at least) snmpd will log all connection attempts, which can turn your /var/log/messages into a giant mess, looking something like this…
Collectl & Colplot Sytem Performance Analysis Tools
It is not often that I run into Unix performance analysis tool that, a), I haven’t seen before and, b), is worth my attention. Collectl is both. It is a very useful combination of iostat, …
Dealing with Runaway Processes
There is no official definition of a “runaway process”. Generally, it is a process that ignores its scheduled priority. It can also be a process that enters an infinite loop. Or it can be a process …
Multi-Cluster Ganglia Configuration
I’ve discussed Ganglia configuration and RHEL/CentOS installation in the past. The usual structure is where every node belongs to one cluster. Recently I ran into a requirement to have nodes assigned to more than one cluster. …
Observium Agent Installation on Pidora
Just some quick notes on installing Observium agent on Raspberry Pi running Pidora. Pretty much the same process applies to any Fedora-like flavor.
Filesystem Performance Testing Using dd
Below is a simple script to test filesystem read/write performance using dd with varying blocksize parameter. This can be useful for testing local filesystems as well as network-mounted filesystems. The end result will be a …
Changing Process CPU Affinity on Linux
A common real-life scenario: on a multi-CPU system Oracle processed have taken over and the system has ground to a crawl. The average system load is in double-digits and even logging in takes several minutes. The possible root causes for the problem can range from inefficient SQL queries (the common problem) to insufficient system resources. But at this point you just need to make the system a bit more responsive, so you can start troubleshooting.
Load-Testing HPC Linux Clusters with “stress”
The “stress” is a simple-to-use load generator for POSIX systems that I found very useful for stress-testing HPC clusters. The current version of the application is 1.0.4 and it was easy to compile and install. Stress can create configurable system load for CPU, memory, I/O, and disks. In the example below we ran “stress” on a SLES 11 HPC cluster with HP CMU 4.2 installed.
Installing Ganglia on RHEL
This is a quick follow-up to my earlier post about installing Ganglia from source on SLES. Here we will install Ganglia from precompiled RPMs on an RHEL server. The basic cluster setup for this example remains the same: two clusters: CLUSTER1 and CLUSTER2 with head nodes head_node1 and head_node2
Copying Data: Are We There Yet?
I am sure this will sound familiar: you are copying a large amount of data – either locally or over the network – and you are wondering how long it will take and if there is a way to make things go faster.You may be surprised, but it does matter what type of files you are copying: 1Gb-worth of many small files will take considerably longer to copy than two 500Mb files. The hardware you are using is an important consideration, but it’s not the only factor limiting data transfer speed.
Testing Filesystem Performance with Bonnie++
Bonnie++ is a benchmark utility designed to test performance of hard drives and filesystems by simulating various types of disk I/O. Bonnie++ may be used to test local disks as well as network-mounted filesystems. It …
Linux and High I/O Wait
When you look at the CPU activity of your computer, one of the parameters is the iowait. This value shows how much time your CPU wastes while it is waiting for I/O operations for complete. …
Linux performance tuning
Linux Performance Tuning
April | May 2007 | by Jaqui Lynch
Note: This is the second article in a two-part series. The first installment was published in the February/March issue.
In last issue’s article I introduced basic Linux* …
Latest Comments