We newly setup Microsoft Azure 1core VM (Region SouthEastAsia) running a website. When we try to access home page from India, it loads after a considerable delay of 3sec to 7sec when there was no load on the server. The delay is consistent, and observed for other pages also.
firebug screen shows waiting time around 1600 ms
Q1: Is this a known network latency issue with Azure ?, if no How can we reduce the latency?
Q2: What is the Microsoft KPI for network latency ?
Environment:
Azure VM: Basic_A1 (1 core, 1.75 GB memory)
Application: LAMP stack (Apache1,php,Mysql)
OS: Ubuntu 14.04 LTS
Region: SouthEastAsia
Accessed from: India
The response size of the home page: 1.4 MB.
Observations:
1. Not a processing delay: We did a wget http://localhost/ on the server, and the page download was done within 700ms
Previously we had same setup with other vendor, the page load took less than 1sec
azure 2 core machine also has the same latency.
I suggest you to first log into an Azure VM and see if you do notice a time latency to the Website. Also you could check from a different ISP and see if the issue persists.
You should try two things:
Check Your site with http://tools.pingdom.com/fpt/ and give us an output (may be as image).
Check performance both with azure host name and Your own host name and compare results.
Related
Every morning when I first login the system, it takes about 13 seconds. And I open the stock item list(1000 items in total), it takes about 16 seconds.
Sometimes these happen when I logout for a while and re-login. Is this normal for Acumatica?
Updates:
I'm trying to preload the whole website through Application Initialization in IIS10.0. But when I config the initializationPage in config editor for Acumatica. I find that it's very difficult to set this. Because the url of Acumatica is changing like below, the number after "W" is changing for no reason. Then I can't set a preload url. Anyone can help on this?
/AcumaticaERP/(W(5))/Main?....
/AcumaticaERP/(W(7))/Main?....
My system info:
OS: Windows 10
Processor: Intel i7 4790K 4Ghz
Memory: 8 GB RAM
Storage: 120 GB SSD
I think you are experiencing what is called 'idle time-out' and this is normal for IIS depending on configuration.
If you are hosting on-premise you can change these settings in IIS (inetmgr.exe). Select the AppPool used for your Acumatica instance, go into Advanced Settings and location idle time-out options:
If the Acumatica instance is hosted by another party like a SAAS provider you should contact them about it. Note that this could affect ressource consumptions for a low traffic website and ressource in SAAS plans could mean money.
For a preload URL you can simply ignore the W(...) part - you won't be able to get there anyway, as it requires authentication. Simply enter the root part (in your case - up to AcumaticaERP/).
We have a Microsoft DC R2 server running only an Interbase database application, all works fine and we can access this application via both Point to Site and Site to Site VPN.
Our transfer speeds for files is coming in at about 5Mbps which is fantastic.
When we access our software (locally) which pulls data from the server (Azure) we're seeing it clock speeds of about 125KBps.
This results in a 3-6 second wait before the dataset appears on screen within our application.
In a local environment this is done within 0.5 seconds.
I'm trying to get to the bottom of the issue as the Internet connection we are on is a 100Mbps feed.
I look at the Draytek router and assumed that this was the problem, however we have tested from multiple sites and ISPs and can't seem to get any improvement on application DB access speeds. SMB speeds remain impressive.
We're not too experienced in the Azure area but we can't work out any way of improving those speeds, if anybody has any suggestions that would be fantastic.
FYI We're using an A2 Windows deployment (approx 4Gb).
Regards,
Pottre11
I have a vm on Windows Azure running centos6 as dev node, and have recently experiencing a huge increase in outgoing traffic.
VM was stopped and after started by 1 day it shows 236.63 GB on outgoing traffic, server not running any applications, only 1 landing page.
need help how to detect what causing the traffic and how to prevent it, since it's reflecting badly on my bill.
I'm want to build a FTP server on Microsoft Azure platform.
The server probably will be based on vsftpd (linux ubuntu server) - not closed. Can select another free ftp service.
I'm have two issues:
End-points - If I'm using Passive mode, I'm need to allocate port range for this. Let's say 8000 to 8100. BUT, I'm have a option to allocate only 20 ports in the "End-points" interface.
I'm need to allocate at least one terabyte of storage in this server. How it's could be done with the machine?
Thank you!
There are a lot of articles written on the subject. Whereas one of the most recent is here.
Fr Virtual Machines, the limit of InputEndpoints is actually 150. Refer to the latest Azure limits compilation here.
As for 1T storage - check the VM sizes for Azure Virtual Mchines. Where A0 (the smallest) supports max 1 data disk of 1TB, and A4, A7 support 16 disks 1TB each (so total of 16TB per VM of size A4 or A7).
As for "built-in" endpoints - you can freely and surely remove them. Especially when you do not use them.
I have an Orchard CMS website currently hosted on Windows Azure Websites.
Its a pretty standard blog where images are hosted via skydrive and linked, so the blog itself only serves html.
I've set it in Shared mode, running 1 instance.
But I keep getting quota reached. and it seems like my site is always maxing out the memory (max is 512mb per hour) and I can't understand why?
I've tried increasing to 3 instances, but it doesn't increase the maximum memory I can use.
Update:
The maximum usage for websites under Shared mode are:
CPU Time: 4 hours per day, 2.5minutes per 5 minute
File System: 1024mb
Memory usage: 512mb per hour
Database: 1024mb (web instance)
Update2:
I've tried re-creating my website in different zones. Currently my site is hosted in US West, which has the above limits, but other zones have slightly different limits, such as East Asia has 1024mb per hour memory usage limit! I haven't been able to dig up any documentation on this, which is puzzling.
Update3:
In Update2 I mentioned that different regions have different "memory usage per hour limit". This is actually not true. I had set up a new site under the "Free" setting with 1024mb per hour, but when I switched this to "Shared" the memory usage limit came down to 512mb per hour.
I have not been able to reproduce this issue in any of my other sites despite being the same source code, which leads me to believe its something weird with my particular azure website set up. Possibly something to do with the dashboard as mentioned by #Vinblad.
I'm planning to set up a new azure website in a different region, and while I'm at it, upgrade to Orchard 1.6
Had a similar issue on Azure with Orchard. It was due to the error log files continually increasing and taking up space. Manually deleting files at the moment but have to look into a more automated solution.
512MB / hour doesn't make any sense at all, I agree with Steve. 512MB (not per hour) is more than enough to host Orchard however. Try to measure memory on your local copy of the site. If you do get abnormal memory consumption, try to profile it and find the module that's responsible for it. If not, then contact Azure support and ask them why the same application would take more memory on Azure than on your local machine.
Another thing to investigate would be caching: do you have output caching enabled?
I saw this post on the Azure forums where they recommend disabling the dynamic module loader. We gave this a try but this gave us problems with the images so we had to revert back.