This post is password protected. To view it please enter your password below:

Password:

Yahoo!'s Exceptional Performance team evangelizes best practices for improving web performance. They conduct research, build tools, write articles and blogs, and speak at conferences. Their best practices center around the rules for high performance web sites.

Rules for High Performance Web Sites

The Exceptional Performance team has identified 13 rules for making web pages fast. Each rule is discussed in the Developer Network Blog articles listed below.

1. Make Fewer HTTP Requests
2. Use a Content Delivery Network
3. Add an Expires Header
4. Gzip Components
5. Put CSS at the Top
6. Move Scripts to the Bottom
7. Avoid CSS Expressions
8. Make JavaScript and CSS External
9. Reduce DNS Lookups
10. Minify JavaScript
11. Avoid Redirects
12. Remove Duplicate Scripts
13. Configure ETags
14. Make Ajax Cacheable

Creation of SVN repository for project on server

1. 1st I installed subversion on my server with apt-get install subversion
2. 2nd I created a directory tree for version control under /usr/local/svn
3. Each project is then contained within its own directory under this i.e. /usr/loca/svn/My_Project
4. I then issue the command "svnadmin create /usr/local/svn/My_Project"
5. Then I import the web site I'm developing into svn with the following: svn import /location_of_files file:///usr/local/svn/My Project/ -m "Initial import"
6. This creates the 1st version in the repository.
7. I then edit the svnserve conf file in the My_Project/conf directory to the following:

[general]anon-access = none

auth-access = write
password-db = userfile

This allows access to the repository via password control.

A userfile is created with the user name=password format

i.e.

[users]

UserName = Password

Accessing Repository

As I do most of my development on a windows machine I installed tortoise SVN http://tortoisesvn.net/ which is an SVN client that is implemented as a Windows shell extension, which makes it integrate seamlessly into the Windows explorer.

Create a Directory where you would like to store a working copy of the web site e.g. c:\web developement\some project

Right click on the directory and select the SVN checkout option on the context senstive menu.

Enter the location of the repository in the subsequent dialog box

This will checkout a version of the web site to the local machine. These file can then be edited using the appropiate editor. Once happy with the files the documents are the checked back in by right clicking the root directory and selecting the SVN comit option.

The typical work cycle looks like this:

Update your working copy

svn update

Make changes

svn add

svn delete

svn copy

svn move

Examine your changes

svn status

svn diff

Possibly undo some changes

svn revert

Resolve Conflicts (Merge Others' Changes)

svn update

svn resolved

Commit your changes

svn commit

Link to subversion manual

Start Slide Show with PicLens Lite

Lean principle is to reduce waste as waste raises costs. Simple things like tidying up the laboratory improves working practices and reduces waste.

Value stream mapping

A process of mapping the flows and movements in the laboratory to identify areas of waste. Simple message is not to try and do everything in one go, leave time for changes to take place and bed in. i.e. split into phases.

Identify bottlenecks 1st and then move on:

e.g.

Reception
then analysis
then courier

Cant impose any changes on staff need to get them to buy into this. Have a small group feeding into a "lean executive" who can make decisions.

Activity of product

(need to get Rachel's data on this and re-audit sample flows within the laboratory)

Map of what sample is doing in time and in space. ? Film the processes to also map what staff are doing. Stress filming the process and not the operator.

Structural layouts drive waste. Don't waste talent, don't install automation before lean principles have been followed otherwise you are automating an inefficient process. Return on investment is typically 15-18 months.

This triggers:

1. Query to find earliest date in table (start_date)
2. Find latest date in table (end_date)

Loop though set table using Setcode

From setcode decide if daily or weekly

query table and count number of samples received in a day.
If > 20 then do daily stats

SELECT Min(s.`DateReceived`) INTO Start_Date FROM setinfo s
ORDER BY s.`DateReceived`;

SELECT Max(s.`DateReceived`) INTO End_Date FROM setinfo s
ORDER BY s.`DateReceived`;

ELSE
do weekly stats

calc_all_stats MySQL stored procedure

DELIMITER $$

DROP PROCEDURE IF EXISTS `activity`.`calc_stats_all_tests_ver2` $$
CREATE DEFINER=`root`@`localhost` PROCEDURE `calc_stats_all_tests_ver2`()
BEGIN

  declare done INT DEFAULT 0;
  declare c_SetName VARCHAR(50);
  declare Start_Date Date;
  declare End_Date Date;
  declare ave_per_day DOUBLE;

DECLARE cur1 CURSOR FOR

SELECT s.`SetCode` FROM setparameters s;

DECLARE CONTINUE HANDLER FOR NOT FOUND SET done = 1;

DROP TABLE IF EXISTS `activity`.`daily_stats`;
CREATE TABLE  `activity`.`daily_stats` (
  `SetName` varchar(50) NOT NULL,
  `HrsIn` double NOT NULL,
  `DateReceived` datetime NOT NULL,
  `AVGTurnaround` double NOT NULL,
  `NumberOfRequests` INT NOT NULL
) ENGINE=InnoDB DEFAULT CHARSET=latin1;

DROP TABLE IF EXISTS `activity`.`weekly_stats`;
  CREATE TABLE  `activity`.`weekly_stats` (
    `SetName` varchar(50) NOT NULL,
    `HrsIn` double NOT NULL,
    `WeekDate` datetime NOT NULL,
    `AVGTurnaround` double NOT NULL,
    `NumberOfRequests` INT NOT NULL
  ) ENGINE=InnoDB DEFAULT CHARSET=latin1;

SELECT Min(s.`DateReceived`) INTO Start_Date FROM setinfo s
ORDER BY s.`DateReceived`;

SELECT Max(s.`DateReceived`) INTO End_Date FROM setinfo s
ORDER BY s.`DateReceived`;

  open cur1;

  repeat
        fetch cur1 into c_SetName;
          if not done then

            SELECT AVG (counts) INTO ave_per_day from(
              SELECT s.`Set`, count(*) as counts FROM setinfo s
              WHERE s.`Set` = c_SetName
              GROUP BY s.`DateReceived`) as f;

            if ave_per_day > 20 then
              CALL calcstats(c_SetName, Start_Date, End_Date);
            else
              CALL calculate_weekly_medians_by_date_range(Start_Date, End_Date, c_SetName);
            end if;

          end if;
  until done end repeat;

close cur1;

END $$

DELIMITER ;

Could be useful in tracking movement of samples through lab in initial modeling phase ? would need to do it all the time maybe just one or two for TAT monitoring etc.

1. Using Excel to chart TATs for Monthly QA meetings. Tables containing daily and weekly TATs are generated and then graphs built from these tables.

2. Web display of data could do as above but on the fly. However as data is unlikely to change on a regular basis (monthly) then tables could be generated as per 1. New data could trigger generation of new tables of statistical data.