Jump to content

sjmiller

Members
  • Posts

    11
  • Joined

  • Last visited

    Never

Everything posted by sjmiller

  1. I am a software engineer with a company, where my responsibility is maintaining and further "developing" a software package called KnowledgeTree for a large customer base. I'm finding that I'm not too fond of this software, as it is extremely intensive on the database, and is poorly coded to increase in database activity based off of how many members are registered through the software. I am in need of a good database optimization for this, and need to configure it, as every ounce of "umph" that I can give it can only help make it anything better than unbearable. The problem is, I am not too spun up on ways to benchmark a pre-built tool that has so many excessive queries being made to the MySQL database. I've benchmarked hand-made applications by going through and pinpointing the problematic database queries and such, but nothing that is already built and ready to run. I would like to be able to pinpoint and optimize the configuration of the database, but I don't really know how to gauge its performance when it is a full fledged application. Are there any tools that would assist me with benchmarking page loads, that I could use as a measure to determine such server variables as the query cache size, or the key buffer size? Essentially, what I think would be optimal is to benchmark the application (specifically, full page loads via URLs), get a baseline set of numbers, then go in and start tweaking the buffers and caches on the database (amongst other things), and see a response on the next benchmark, helping me determine what would be most optimal. I've attempted to use Apache's benchmark tool, but I'm not noticing any fluctuation in performance when I go to the extreme with things like query caching (albeit with caching, it would only be noticeable on the first and second page load), or key buffers, limiting them to almost nothing, then blowing them wide open. What other tools would be of use in a test like this? Thanks for your advice.
  2. Is there any way to generate a select form with options deriving off of the values of an ENUM field in MySQL? I'd like to cut out the middle man of having to manually go in and type out the select options so that they match the values of the enum field, and if possible (long shot here) be able to easily edit the values in the ENUM via a form. Is this a little extreme or doable?
  3. Oh wow, was unaware that I was able to do an 'as' turning the associative array into two separate variables! That's a neat trick! Thanks!
  4. So I have what could be an extensive form that requires users to fill out all form inputs. If even one of these inputs is missing, I need to have it return an error. I find it somewhat inconvenient to go through 20-some isset()'s or empty()'s in one if() statement. Is there any easy way to just count the $_POST array's variables that contain data without having to make an extensive one-by-one validation that each $_POST value has been filled out? Thanks ahead of time for your help!
  5. It's a very smart concept that is very efficient with regards to file sizes and request loads (as haku mentioned) It's a very simple concept as well that you can use with CSS (yes, it is CSS). One way to try and re-emulate such an effect is to try this is your code .class { width: 40px; height: 40px; background: url("some/image.jpg") no-repeat 20px 20px; } All that does is it take the class named "class", which has a width of 40px and a height of 40px, then paste a background image (named image.jpg), which starts (or is anchored) 40px down the to the right, and 40px down the image (0px 0px is default, which anchors to the top left of the image, not the bottom). So anything that is above and to the left of 40px in the image will not get displayed. A simple way to explain it visually would be like such: say this was an image +-------------------+ | | | | | | | image | | | | | | | +-------------------+ Now if we call this image and tack a background position of 10px 10px, it will only display what is 10px from the left, and 10px down of the image, so essentially, this is what you'll see of the image: +-------------------+ | not displayed | | +--------------+ | | | | | | | | displayed | | | | | | | +----+--------------+ That's just a theoretical example, saying that to anchor point of the displayed portion was 10px down and to the right, but you should get the point. So say in your image, you have 4 images within that image, each taking up a quarter (2 images in the top portion, 2 in the bottom), each image being exactly identical size wise, and all take up exactly a quarter of the image: +---------+---------+ | | | | 1 | 2 | | | | +---------+---------+ | | | | 3 | 4 | | | | +---------+---------+ you could even specify 50% 50% for part 4 of the image or 0 50% for part 3 or %50 0 for part 2, (0 0 being part 1) Hope this clarifies how that works!
  6. I want to re-echo this. This is without a doubt the best route to take for beginners. I think the toughest leap for new beginners that will help make the learning curve smoother is to understand how exactly things like MySQL work, or how the database is structured. Upon first examining the concepts of a database, all one can assume is that it is a program that runs somewhere and holds something, somehow. To understand that MySQL works in un-ordered rows of "textual" data is a huge leap in understanding how one would utilize it. Much like OOP, if you don't understand the concepts behind how it works, in a general (not syntax necessarily) sense, you may want to start with the basics of concepts, and then start learning syntax, and how to script with it. Your application that you are describing is a great start in all reality, the functionality is basic, and as Mchl stated, you are going to learn a lot by doing it yourself for this first time, and once you learn everything, you are going to want to start over (guaranteed, I'm in the middle of re-doing some of my older projects because of how much I learned while developing them). Use this as a prime opportunity to learn as much as you can, and you'll find out just how easy it is to accomplish what you want.
  7. No worries, I installed and applied the firefox add-on listed above, and it works wonders, with the short downfall of excluding localhost. I can see being more of a pain to implement than it is probably worth. But otherwise, it works wonderfully, and it accomplishes what I am looking for! Solved! Thanks for the help!
  8. Ah, so the reality of web development settles in. I've encountered many similar situations. How well commented is the code? How well structured is it? Is it uniform across the board with naming conventions, bracket placement, indentation? When re-working (or completely redoing) these are the things that you have to take into account when you could possibly have to pass it off to someone else. Make sure that you are continually maintaining your code and making it legible and seeing that it flows smoothly. 1, it's easier for you to go through and find errors, and 2, others don't have a heart-attack when viewing it.
  9. I am sorry for not clarifying. Out in Iraq we function off of a MS IIS server. Back here in the states, I intend on using apache for testing, as I have no intention of dealing with the hassles I've experienced with IIS. Apache for development and testing, IIS for the real deal. With that said, any chance you could link a couple of these modules? I wouldn't know where to go to find them. Thank you very much!
  10. Greetings all, being that this is my first post on the boards, I hope this question is in the right section. Reading over the forum rules, I believe this to be right location. My situation is as follows. I returned from another successful rotation out to Iraq this fall. While out there, I had a collateral duty involving maintaining my battalion's NSAnet website. While "maintaining" this server (Microsoft IIS v.?) I single-handedly developed a LARGE scale php & mysql application to aid in the development of intelligence while out there. Unfortunately, it was across a satellite connection, and we had sites all over the place. Traveling from satellite to satellite proved painful for the size of this application, and most users never got the full experience because they figured that the use of the application did not outweigh up to (and sometimes exceeding) 30 seconds for a single page load. I attempted many resolutions to fix this problem, to include cutting down the CSS, trying to condense the mysql queries, and optimize the db with proper uses of indexes, etc.. etc.. But I was still facing incredibly slow load times. Now that I am volunteering to go back out in Feb, I want to re-construct the application before I go back, and take advantage of the time I have. With re-construction, I want to over-emphasize on swift load times to combat the killer of my application. In order to properly test this, I want to emulate the scenario as much as possible. I am looking to try and throttle my connection for testing purposes to (no joke) under 5kb/s load times. I was told that some routers have a QoS service on them that would allow for this, unfortunately I don't have that functionality. Does anyone know of any software that would assist me in throttling my connection for testing purposes to a set rate that will not exceed or fluctuate? Thank you for your help!
×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue.