A client require SLA that guarantees usage of an X degree of concurrent users. Do you find it normal to offer guarantees on concurrent clients in SLA in case yes, how to calculate concurrent users
Concurrent consumers for what
My respond to without knowing much info would be. Make sure ones hardware is spec’d to take care of that many users. Also make sure your customer is paying for this particular SLA.
For the web app (PHP/MySQL).
Make sure your own hardware is spec’d along with you’re charging the perfect amount to guarantee that SLA. It should be as simple while that.
Granted I have no idea what you’re loading, what it’s with regard to, what hardware it really is running on, what it can, etc.
It ought not matter what it really is, if they desire that SLA; fee accordingly.
I’m performing a fantasy football online app for client. How much regarded as a charge for SLA pertaining to concurrent users. And learn how to measure the concurrent users my server (LAMP) could support
My LAMPs
-1 apache machine (4 core 4GHZ, FOUR GB) and ONE PARTICULAR mySQL server (4 Center 4GHZ, 7GB)
Absolutely definitely not. You have no control in the programming of the customer, or the traffic being received by the server, abusive or otherwise, or anything else that could be important for having these kinds of guarantee. If it is a dedicated server, you’ll be able to guarantee they’ll obtain server they’ve called for and that the idea remains up plus functional. So many things can not work right with a regular php / mysql / light setup that there is not any reasonable way that will guarantee any particular quantity of simultaneous users.
edit: you’ll be able to just pick lots and then charge for this and consider the idea like insurance. You already know, like, calling Auto insurance an “SLA” towards crashing your vehicle. Obviously eventually it’ll happen and it’s totally beyond the control in the car company, but they can easily still charge a premium for it on the off chance they’ll should pay out.
Quote: I’m performing a fantasy football world wide web app for purchaser. How much regarded as a charge for SLA pertaining to concurrent users. And learn how to measure the concurrent users my server (LAMP) could support
My LAMPs
-1 apache machine (4 core 4GHZ, FOUR GB) and ONE PARTICULAR mySQL server (4 Center 4GHZ, 7GB)A lot is based the server configuration and also the website.
You could begin with testing it with something like ab or Load Impact for any real world solution.
Quote: A client require SLA that guarantees usage of an X degree of concurrent users. Do you find it normal to offer guarantees on concurrent clients in SLA in case yes, how to measure concurrent usersNo, it can be completely not natural.
But, I figure you can work out this numbers after adequate load testing plus a heavy premium.
If you might be offering an end to end managed service then it really is reasonable for the customer to have an expectation of the quantity of concurrent users may be supported, although this needs to be defined as part of the acceptance criteria without the other method around. Selenium is a great industry standard automation tool which may also be used for that purpose.
You can easily always add a clause to you service agreement which excludes things you might have no control over (or the location where the client is not prepared to afford mitigation) such seeing that DDOS, Hardware Inability etc.
I would check with your legal suggest regarding this. Guaranteeing fitness for any particular application (especially one you cannot control) may open you up to an array of nasty liability challenges.
Quote: I’m performing a fantasy football world wide web app for purchaser. How much regarded as a charge for SLA pertaining to concurrent users. And learn how to measure the concurrent users my server (LAMP) could support
My LAMPs
-1 apache machine (4 core 4GHZ, FOUR GB) and ONE PARTICULAR mySQL server (4 Center 4GHZ, 7GB)It’s not unusual for managed services to get provided on your per-mailbox, or per-desk, or per-transaction basis and on the limited description we have, this sounds not very much different.
If you control the complete end-to-end solution (application, national infrastructure, servers, network etc) then I don’t even think it unreasonable for any client to ask to take the risk about the hardware being inadequate.
If the required forms were the customer’s, then that is often a completely different report…..
As some have said, just make sure the liability clauses are usually clear and rational.
Quote: Absolutely not. You have no control in the programming of the customer, or the traffic being received by the server, abusive or otherwise, or anything else that could be important for having these kinds of guarantee. If it is a dedicated server, you’ll be able to guarantee they’ll obtain server they’ve called for and that the idea remains up plus functional. So many things can not work right with a regular php / mysql / light setup that there is not any reasonable way that will guarantee any particular quantity of simultaneous users.
edit: you’ll be able to just pick lots and then charge for this and consider the idea like insurance. You already know, like, calling Auto insurance an “SLA” towards crashing your vehicle. Obviously eventually it’ll happen and it’s totally beyond the control in the car company, but they can easily still charge a premium for it on the off chance they’ll should pay out. Guys he said it can be his web job application.
It really is his application thus he should absolutely be able to predict performance immediately after sufficient testing certainly.
We helpful to use QALoad for such type of testing. Of course i not able to tell you what it should be regardless of the specs you give us because it is your application. You simply must test using something like QALoad or maybe getting a lot of testers and calculating performance.
thx folks. You all tend to be awesome.