Really dedicated resources? (request timing)

Options
Vladimir Slavinsky
Vladimir Slavinsky Member
edited February 2023 in ❓Other questions

Hello

Today I am glad that I became a client of Xano (launch package, Germany region). The frontend is not yet available to users and today I wanted to feel at least some difference in the timing of request processing, but I found some strange delays.

Is this normal behavior?

What packages will not have such delays?

Thank you

Best Answer

  • Sean Montgomery
    Sean Montgomery Administrator

    ADMIN

    Answer ✓
    Options

    @Vladimir Slavinsky A lot of things take place when processing lots of statements. The additional latency could be the database taking a bit longer to index something. It is hard to tell without looking into it in more detail.

    In this particular example, your 800% difference happens to still be under a tenth of a second, so the total outcome is still very fast. We definitely want to keep things as fast as possible, but in this particular example, I wouldn't worry about it. If you had something taking multiple seconds, then that would be more interesting because it would easier to track down the underlying reason.

    The Scale plan is significantly faster than the Launch plan, and the Scale+ plans keep getting faster. We provide more details into our plans in this documentation.

Answers

  • Michael Udinski
    Michael Udinski Administrator

    ADMIN

    Options

    Hi @Vladimir Slavinsky while performance is dependent on numerous amount of factors including the CPU, server location vicinity, etc. Launch plan is the introductory dedicated resource plan, meaning it will have the lowest resources of all paid plans. That being said, your loop of 20 operations still seems quite fast. I'm curious about what type of timing you are expecting for this operation?

  • Vladimir Slavinsky
    Vladimir Slavinsky Member
    edited February 2023
    Options

    @Michael Udinski

    It is clear that these are the lowest resources, given that this is the cheapest package.

    I do not expect any specific timings, I expect that when 1 user at the same time makes a similar request, then all operations in the cycle will have a similar timing (well, at least + - 50%). I dont know (dont understand) why the timings of the same operations in the loop have a difference of up to 800% (as at screenshot). Here is my question.

    And one more question, how many more resources will I get in the next plan (Scale x1)? How much it will be growth perfomance compared to Launch?

    Thank you

  • Vladimir Slavinsky
    Options

    @Sean Montgomery

    Thanks for the answer.

    This moment is important for me, because it is assumed that each user will add about 500 such records at a time once or twice per month. Now I have about 6-8 seconds to complete this stack. At the moment, this is not critical, while I still have few users, however, I would be interested in how long it took to complete such a stack with the Scale x1 plan.