31
votes
How to measure req/sec by analyzing apache logs
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 ...
- 491
14
votes
How dangerous might it be - and what performance gains may be had - by turning vulnerability mitigations off on non-Internet facing servers?
The Linux kernel flag mitigations=[on|off] is a single toggle to easily enable/disable all available kernel mitigations for hardware vulnerabilities as listed here https://docs.kernel.org/admin-guide/...
- 1,899
11
votes
Options for performance improvements on very big Filesystems and high IOWAIT
I have a similar (albeit smaller) setup, with 12x 2TB disks in a RAID6 array, used for the very same purpose (rsnapshot backup server).
First, it is perfectly normal for du -hs to take so much time ...
- 51.3k
9
votes
Ansible performance issues on a huge group
Increase forks
You have significant memory and CPU, so a few hundred forks aka worker threads is reasonable even if it they are heavy on resources. ansible.cfg:
[defaults]
forks = 350
serial is the ...
- 34.2k
7
votes
Accepted
Disadvantages of using ZFS recordsize 16k instead of 128k
Short answer: It really depends on your expected use case. As a general rule, the default 128K recordsize is a good choice on mechanical disks (where access latency is dominated by seek time + ...
- 51.3k
6
votes
Options for performance improvements on very big Filesystems and high IOWAIT
This Filesystem stores a huge amount of small files with very many SEEK operations but low IO throughput.
🎉
This is thing that catches lots of people nowadays. Alas, conventional FSes do not scale ...
- 9,615
6
votes
Packet drop on HP ProLiant DL360 G9 running RHEL 6.10
Check the driver version. There was the bug with correct accounting of the rx_no_dma_resources, when the rx buffer is full. So check the length of the ring buffers (ethtool -g <iface>) and ...
- 5,302
5
votes
nfs server getting many rpc-srv/tcp: nfsd: sent only 700260 when sending 1048708 bytes - shutting down socket
This site reports some more suggestions, including increasing the NIC txqueuelen. These suggestions are mainly aimed at high latency networks, however I noticed that under Fedora24 Infiniband ...
- 151
5
votes
Stand alone single thread optimized server vs vm running on VMWare
You are correct in your assumptions. For your particular case stacking Mhz's instead of vCores is a right way to go. Take a look at this processor: (features 4x 3,5-3,7 Ghz physical cores + HT).
...
- 5,135
5
votes
Serving a static website with nginx. Response time over 600 milliseconds. What’s wrong?
I re-ran the test from Mumbai with no bandwidth restriction and got a TTFB of 237ms for the page and 7ms for a static resource.
You restricted the test to 2G, which is very low bandwidth.
So the ...
- 33.2k
4
votes
Will noatime on a tmpfs volume improve performance?
In theory I agree with what @kasperd said. However, I decided to try the change on my very busy tmpfs backing munin data, and it immediately provided an 8% decrease in total munin processing time.
...
- 1,053
4
votes
Find out which task is generating a lot of context switches on linux
In newer kernel version
sudo perf record -e context-switches -a # record the events
# then ctrl+c
sudo perf report # inspect the result
This will give you the exactly result about context-...
- 141
4
votes
Accepted
Dual-port RDMA Throughput Issue
I think the bottleneck here is the PCIe link between the ConnectX and the host. The ConnectX-3 has a Gen. 3 x8 PCIe connection, which is bound to a theoretical maximum of 63.04 Gbps (according to this ...
- 272
4
votes
Accepted
How do I tune a single-threaded server to be amazingly fast?
Providung a laundry list of tunable parameters here will do little to put your server into a better state. Performance tuning is a process, not a task. Measure, tweak, and measure again. Restrict ...
- 326
4
votes
php memory_limit vs kubernetes resource memory limit
Setting up container resource requests and limits is the first step towards using resources effectively in your Kubernetes cluster. After you set them, make sure you have monitoring and alerting in ...
- 465
4
votes
Accepted
Is it possible for Windows clients to carry out external queries through an alternative DNS server directly?
No, the conditional forwarder are configured on the DNS server side, not on the client side sadly
I would deleguate a local server with a DNS role that could redirect to the remote DC all your domain ...
- 17.3k
4
votes
Accepted
MySQL LOAD DATA INFILE: Better Server, Worse Performance
Here's what seems to have been causing this behavior:
Per the Azure documentation, the Basic tier server on Azure comes with "variable" IOPS whereas the Memory Optimized server comes with a fixed ...
- 71
4
votes
Disadvantages of using ZFS recordsize 16k instead of 128k
I recommend tuning if and when you encounter a problem.
ZFS defaults to 128K record sizes, and that is acceptable and valid for most configurations and applications.
Exceptions to this include:
...
- 199k
3
votes
Hyper-V Windows Server 2016 VM is super slow
Slow startup times are usually connected with slow storage performance and depending on the size of virtual machine disk may take more time to start completely, however, 1 hour is "a bit" too much. I ...
- 6,309
3
votes
What performance statistics can I use to determine if hyperthreading will improve performance?
I can't tell if you're too academic in approach, but for most workloads on modern CPUs, people recommend leaving hypertheading on.
Now, given that you haven't provided details like the application ...
- 199k
3
votes
Accepted
Debugging and fine-tuning the linux process scheduler
Is it correct that the php-fpm process spent as much as 58.130 ms waiting for a CPU to run on?
it's time from wake-up to context switch, so I'd say it's time taken to make it finally running (thus — "...
- 9,615
3
votes
Is there is any tool/way to find out why packet is getting dropped due to an arp cache full?
There is several articles about ARP table overflow. This article have good explanation. Also, you could check this bug.
Both articles say that you should see error neighbour: arp_cache: neighbor ...
- 4,648
3
votes
Tune Linux & Nignx to handle 10k Connections @10Gbps Server
Already answered by Brandon. Turn on irqbalance. Run numad and tuned. Stop trying to tune unless you have a specific workload that requires it. Where are your wrk test results from testing 2000-10000 ...
- 31
3
votes
Disadvantages of using ZFS recordsize 16k instead of 128k
For what it's worth, it's a recommendation to set "recordsize=16K "according to zfs'documentation itself.
https://openzfs.github.io/openzfs-docs/Performance%20and%20Tuning/Workload%20Tuning....
- 135
3
votes
Ansible performance issues on a huge group
You should try using Mitogen for Ansible, which replaces the host communication part in Ansible with a different approach, and by their words, increases execution speed 1.25x - 7x and decreases CPU ...
- 38k
2
votes
Accepted
How to see actual size of tcp send and receive buffers?
Direct answer to this question: run ss -im dst destination.ipv4.address.here and check skmem rb value (full receive buffer size) and rcv_space value (receiving window). The application buffer size is ...
- 361
2
votes
Does 'tuned' need to keep running?
As of CentOS 7.2 tuned now has a no-daemon mode which can be turned on by setting daemon = 0 in /etc/tuned/tuned-main.conf. This is mentioned in the RedHat Performance Tuning Guide.
- 121
2
votes
What does the following C-State output mean?
The standard solution for anything that requires low-latency or deterministic processing/response times is to disable C-states entirely.
I don't know what type of servers you're using, but HP has a ...
- 199k
2
votes
Accepted
Tuning an email backup dataset on ZFS
I ended up doing what @ewwhite said and did a quick benchmark. My conclusion is that 128k is fine.
BlockSize CompRatio du-s
1 0 50747244
2 1 26001757
4 ...
- 186
2
votes
Two identical systems with very different performance
My recommendations today with EL6 systems on enterprise hardware are the following:
Set your Dell servers to "OS Control" mode for power, versus a "High Performance" or "Dynamic" mode. This will ...
- 199k
Only top scored, non community-wiki answers of a minimum length are eligible