94

I suggest that you make the adjustments they have requested. Then benchmark the performance to show them that it made no difference. You could even go so far to benchmark it with LESS memory and vCPU to make your point. Also, "We're paying you to support the software with actual solutions, not guesswork."


67

Providing you are confident you are within the given system specs they document. Then any claim they are making in regards to requiring more RAM or CPU they should be able to back up. As the experts in their system I hold people to account on this. Ask them specifics. What information provided on the system indicates more RAM is needed and how did you ...


21

This great article helped me a lot... http://www.inmotionhosting.com/support/website/server-usage/view-level-of-traffic-with-apache-access-log I had created a set of prepered commands that I am using to analyze apache log: request per hour cat access.log | cut -d[ -f2 | cut -d] -f1 | awk -F: '{print $2":00"}' | sort -n | uniq -c request per hour by ...


21

ISPs often prioritize traffic to speedtest.net so that they can brag how fast their connections are, while in reality, they don't provide that much bandwidth. They're perfectly aware that most users will only check that site for confirmation. You also have to keep in mind that transfer speed relies both on the client and the server. In today's world most ...


17

The big thing is to be able to prove that you are using best practices for your system allocation, notably RAM and CPU reservations for your SQL server. All this being said the easiest thing is to make the adjustments requested, at least temporarily. If nothing else it tends to get vendors over feet dragging. I can't count the number of times I've needed ...


17

For this specific situation (where you have VMware and application developers or a third party who does not understand resource allocation), I use a week's worth of metrics obtained from vCenter Operations Manager (vCops - download a demo if needed) to pinpoint the real constraints, bottlenecks and sizing requirements of the application's VM(s). Sometimes, ...


17

In addition to the other reasons posted, TCP connections don't work well with large files when the bandwidth-delay product becomes large. Like on an otherwise fast connection to an island. See Wikipedia's entry on TCP tuning. So Speedtest can dump a small file through the connection at 95 mb/sec, but wget can only get 10 mb/sec on a 20 MB file.


15

A machine is idle when it's not performing the task it's supposed to perform for lack of requests. For example, if you had an email server, you could determine if it was fielding any requests from the email application on it. If it isn't, and that's the only thing on there, then it's idle. Things of course get more complicated with collocated services on a ...


12

If you want to see each file as it is being opened, you can filter that with strace. For example: strace -p _pid_of_app_ -e trace=open,close


11

By default Windows hide the entry because of a noted significant performance impact in collecting Disk metrics on a Server: You need to register the Disk Perfcounter before you see this entry. close Taskmanager open a cmd.exe as admin run diskperf -Y close cmd.exe and run Taskmgr Now you see the entry: To analyze disk activity deeper, install the WPT, ...


11

See the sshping utility: https://github.com/spook/sshping Example: # sshping 172.16.47.143 --- Login: 1725 msec --- Minimum Latency: 4046 nsec --- Median Latency: 11026 nsec +/- 0 std dev --- Average Latency: 178105 nsec --- Maximum Latency: 8584886 nsec --- Echo count: 1000 Bytes --- Transfer Speed: 11694919 Bytes/second # sshping --help Usage: ...


11

On Ubuntu 16.04 I just ran into this same issue. After some grousing around I found I needed to enable the service by editing the file /etc/default/sysstat Change ENABLED="false" to true: # Should sadc collect system activity informations? Valid values # are "true" and "false". Please do not put other values, they # will be overwritten by debconf! ENABLED="...


10

I used to work in support - and part of what you're asking sounds highly rational (and probably is): but there are a few questions to ask yourself prior to just doing the "performance enhancement" they're requesting are you running at least at the vendor's stated minimum system requirements already? if you're at least at minimum sysreqs, are you ...


9

Comparing the number of CPUs of a mainframe image isn't meaningful, most likely. Mainframes are designed to run multiple tasks at the same time, and give priority to whatever the business says is most important, and are capable of being very heavily virtualized, so looking at the number of CPUs doesn't tell you much. You have to understand the environment ...


9

Kevin mentions a number of great points, but it might help to start the analysis at a little higher level: what are the two machines and, since we're talking about Tomcat which runs in a JVM, do both have zAAPs or zIIPs (assuming zAAP on zIIP)? From the "d m=cpu" you should be able to get the machine model information which will at least let you know if you'...


8

New Relic says themselves that they use the .Net profiling API. https://docs.newrelic.com/docs/dotnet/new-relic-net-installation#trouble The .NET agent instruments application and system code using BCI (byte code injection). This is done by registering the New Relic agent with the CLR (Common Language Runtime) as a "profiler." The CLR calls out to ...


8

Either ask to escalate the ticket or ask for a different rep. Depending on which vendor it is escalation may help if you say that you feel that the current level of support doesn't adequately address the issue. If they will not escalate then asking for a different rep may help because that requires much less "justification" since all it needs is to not be ...


8

Was trying to do this myself and came up with this. Probably there is a simpler way, but this is what I came up with. First, prepare pipes which will be used to make the benchmarking program communicate through the SSH connection. $ mkfifo /tmp/up /tmp/down Then establish a connection in ControlMaster mode without executing any remote command. This allows ...


8

I skipped some steps suggested by @nicht-verstehen: python -m timeit --setup 'import subprocess; p = subprocess.Popen(["ssh", "user@host", "cat"], stdin=subprocess.PIPE, stdout=subprocess.PIPE, bufsize=0)' 'p.stdin.write(b"z"); assert p.stdout.read(1) == b"z"' Where python -m timeit executes the timeit Python module. The -s/--setup option tells timeit ...


8

"Idle" is not a black and white concept: even a busy server can have free CPU cycles, IOPs and memory to run some other application. As a rule of thumb, target CPU load should be at around ~80% and not above ~90%, as maxing out the CPU will greatly increase system latency. CPU load under ~60% generally meas your server is underutilized. Please also ...


7

Looking at top output is completely wrong. It's about the IOPS. To get a view on the NFS statistics, use nfsstat: Server rpc stats: calls badcalls badauth badclnt xdrcall 40833255 0 0 0 0 Server nfs v3: null getattr setattr lookup access readlink 0 0% 1411374 3% ...


7

The tload command represents the pictorial representation of the average system load through ASCII graph. This command can be used to provide the graph on a terminal. The syntax for the command is: tload [options] [terminal] If terminal is not provided as the argument to this command, then by default it outputs the graph on the current terminal. So the ...


7

The sysstat service and associated timers need to be started: systemctl start sysstat sysstat-collect.timer sysstat-summary.timer To auto-start on next system startup: systemctl enable sysstat sysstat-collect.timer sysstat-summary.timer


7

wget give good practical measure of the speed. The tests of Speedtest probably include kind of parallelism which can explain higher numbers. For good average speed test I think the time for download should be at least 90-120 seconds (to get good average)


6

You could try sar grapher at http://www.fsprings.com/neat/sargrapher.html, you upload sar -p -A output on it provides a page with graphs. If you want you can select the sar options you want it will only graph those.


6

Did ESXi 6.0 completely removed the performance data for VMXNET3 or is there is a way, how can enable throughput logging/graphing for VMXNET3? No, it's just there in vCenter - if you're not seeing you have an issue, you may need a reinstall. EDIT - After playing around all day I have to apologise, you're absolutely correct, I don't know what I was ...


6

Typically you get what you pay for and relatively cheap hosting providers can only offer their services at a competitive price by overcommitting CPU several times i.e. on a server with 48 cores and 128 GB of memory they can then try to host potentially 100+ 1-core/1GB-memory VPS instances. As long as most VPS's are mostly idle, that works well enough, ...


5

Why is "Processor load is too high" false alarm in your case? It's real symptom for me - CPU is saturated. IMHO: use only {Template OS Linux:system.cpu.load[percpu,avg1].avg(5m)}>5 but threshold depends on your server - what and how is it doing. But >5 value is suspicious for me. Example: CPU usage can be low, but CPU load high - in this case it can ...


5

Of course it would. Why wouldn't it? There's nothing saying you can't use both perfmon and vCenter counters to monitor the performance of your VMs - perfmon is better at or easier to set up for some metrics, and vCenter's counters are for other metrics, so it makes sense to use whichever one fits better for the specific metric and purpose you're monitoring....


4

The key difference between the New Relic .NET Agent and traditional profiling tools is that the New Relic .NET Agent doesn't instrument everything. Developers at New Relic have gone through and picked out key methods in various frameworks (e.g., ASP.NET) and libraries (e.g., SQL Server) and the agent injects code into those methods only. A traditional ...


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