Tag Info

Hot answers tagged

24

CPU time is allocated in discrete time slices (ticks). For a certain number of time slices, the cpu is busy, other times it is not (which is represented by the idle process). In the picture below the CPU is busy for 6 of the 10 CPU slices. 6/10 = .60 = 60% of busy time (and there would therefore be 40% idle time). A percentage is defined as "a number or ...


10

The CPU time is the time that the process is using the CPU - converting it to a percentage is done by dividing by the amount of real time that's passed. So, if I have a process that uses 1 second of CPU time over a period of 2 seconds, it's using 50% of a CPU. In the case of your MATLAB process, 217% indicates that it's used 2.17 seconds of CPU time per ...


7

There are many limits for transferring many small files. Some have already been mentioned: network latency, disk write speed, etc. However most of those can be optimized best by using "rsync". If the files don't exist on the destination, and you are pretty sure the process won't be interrupted, using tar piped to tar will be very efficient: cd /SOURCE/DIR ...


3

Brandan Gregg has made a number of visual diagrams of which tool to use on what part of your system that make really nice cheat sheets: Tools: Which options to use in sar to monitor: And closely related: Translate observations into tuning with:


2

They are unsigned long integers with a length depending on the platform (32/64 bit). Yes, they can indeed wrap over if the system is very active for very long. You have to consider that when doing any kind of monitoring. See https://www.kernel.org/doc/Documentation/iostats.txt


2

Based on the documentation you quoted I find it quite clear that one is covering the entire duration from one system call to the next, while the other covers only the time within a system call. The percentage of time spent inside system calls versus the percentage of time spent outside system calls will roughly tell you if a process is CPU bound. A CPU ...


2

The time outside the system call is the time spent running your program's code before it gets to the next system call.


2

DNS operates best in a Master/Slave scenario. Consider this: you have two nameservers, ns1 and ns2. To help balance the load, you have two NS records in your domain record for both ns1 and ns2. By the very nature of DNS, clients will query the nameservers in random order because nameserver lookups are returned in a random order. See this answer for helpful ...


2

A JBOD (Just a Bunch Of Disks) is just that. It has no controller aside from what is necessary to provide power and data lanes to the disks. In a case of a more sophisticated JBOD, you will see a backplane and, depending on the technology used, maybe an expander. The controller needs to be placed on the host's side, it will be connected to the JBOD with ...


2

If you use the latest stable version of proftpd (1.3.5) then there is bug when using TLS: http://bugs.proftpd.org/show_bug.cgi?id=4108 TJ Saunders 2014-09-24 02:46:33 UTC Pull request opened which fixes the issue: https://github.com/proftpd/proftpd/pull/48 Short version: we do NOT want the TCP_CORK/TCP_NOPUSH socket option enabled on ...


2

There is no clear cut answer to such question, however I will provide you with some links that should set you on the right path of optimizing the VM setup. I would start by reading the official documentation from Microsoft regarding performance: Performance Guidance for SQL Server in Azure Virtual Machines: ...


2

The first thing you need to look at is to determine where your bottleneck is. Is the system's CPU pegged, or is there CPU available but the web application's processes aren't using it effectively (ie, not enough worker threads)? Is the system out of RAM, or doing a lot of swapping? Are you bottlenecking on your database? Do you have slow queries, need ...


1

Look into the buffer command, basically it speeds allows simultaneous read/write activity instead of the normal read, then write, then read, etc.... Basically it does this by spawning two sub-processes. The processes communicate using a shared memory buffer. The command line parameters are similar to dd with the addition of parameters to size the shared ...


1

Look into using background jobs for the Get-ScheduledTask cmdlet. This will run the command asynchronously. foreach ($computer in $computers) { start-job -scriptblock { #Parse Args $CompName = $args[0] $TaskName = $args[1] Get-ScheduledTask -TaskName $TaskName -CimSession (New-CimSession -ComputerName $CompName) } ...


1

JBOD used generally means exactly what it says on the tin - just a bunch of disks. There's no raid, no disk consolidation, nothing. You'll see a whole bunch of separate devices down your controller. Intel's 'JBOD' unit needs a RAID adaptor if you want to do anything more clever. For example: ...


1

CPL is load average figures reflecting the number of threads that are available to run on a CPU (i.e. part of the runqueue) or that are waiting for disk I/O. You seem to have ~16 processes that seems to be waiting for the disk. That's the reason you see the cpu mostly idle, it doesn't have anything to do other than waiting for the disk. I would check the ...


1

Keep in mind that harddrives have a certain sector size. Writing normally goes per sector, so you should expect lower speeds below the sector size. In additionn most raid configuration either apply mirroring (same data on multiple disks, speeeds up reading as information only needs to be read from 1 disk instead of X disks, however writes have to go to all ...


1

There is an elegant solution developed by William Glick: parallelizing rsync. /bin/bash # SETUP OPTIONS export SRCDIR="/folder/path" export DESTDIR="/folder2/path" export THREADS="8" # RSYNC TOP LEVEL FILES AND DIRECTORY STRUCTURE rsync -lptgoDvzd $SRCDIR/ /$DESTDIR/ # FIND ALL FILES AND PASS THEM TO MULTIPLE RSYNC PROCESSES cd $SRCDIR; find . -type f | ...


1

Removing unused modules sometimes has the opposite effect as the server is then not able to cache effectively for example, but assuming this is not the problem, the other issue I think might be happening is that you may be low on memory and therefore the so called "disk thrashing" (high IO) occurs, this is very possible specially on a virtual host. A ...



Only top voted, non community-wiki answers of a minimum length are eligible