At my real job, we monitor and report on the performance of a number of different e-commerce verticals.
[I will now try and stop using marketing-speak.]
In the retail index, the Gap has been down for nearly 3 weeks. Well, they have come part way up, but it is flaky, and they don’t have a search function.
No search function? Oh yeah, let’s go shop somewhere else.
Anyway, Erick Schonfeld at Business 2.0 just linked to a story at adfreak about this little re-design outage.
20 days. And counting.
This is unheard of in the Internet economy. Well, not unheard of, but the only other site I know that had a similar re-design outage was Wal-Mart back in 2000.
The Gap is a savvy online retailer. Why couldn’t this been done in a more intelligent way? What has made this re-design so problematic?
Month: September 2005
GrabPERF: Search Index Weekly Results (Sep 12-18, 2005)
The weekly GrabPERF Search Index Results are in.
This Week’s Notes
- ERTW.com Agent location taken offline
- Addition of Google Blogsearch to the Index
- Numerous performance improvements to the GrabPERF interface
Week of September 12-18, 2005
TEST RESULT SUCCESS ATTEMPTS -------------------------- --------- ------- -------- PubSub - Search 0.2688096 99.82 6532 Google - Search 0.4013164 99.97 6532 Google Blogsearch - Search 0.5818507 98.60 4214 MSN - Search 0.6981630 99.83 6532 Yahoo - Search 0.7159974 99.95 6527 eBay - Search 0.8345692 100.00 6528 BlogLines - Search 1.0204595 99.95 6531 BestBuy.com - Search 1.1687228 99.97 6530 Feedster - Search 1.3112797 99.82 6531 Technorati - Search 1.3240335 99.95 6528 Amazon - Search 1.5195445 99.72 2481 Newsgator - Search 1.5823492 99.72 6529 Blogdigger - Search 1.7142475 99.97 6506 BENCHMARK RESULTS 2.0313721 99.50 76849 IceRocket - Search 4.2792600 98.79 6515 Blogpulse - Search 6.5226776 99.29 6522
These results are based on data gathered from two remote measurement locations in North America. Each location takes a measurement approximately once every five minutes.
The measurements are for the base HTML document only. No images or referenced files are included.
Technorati: GrabPERF, Web Performance, Web Performance Measurement, Web Performance Monitoring, search index, SEO
IceRocket: GrabPERF, Web Performance, Web Performance Measurement, Web Performance Monitoring, search index, SEO
Performance Improvements to b2evolution
I upgraded b2evolution yesterday to 0.9.1, which the developers claim had substantial performance improvements.
I believe them.
Kudos to the b2evolution team!
Songs That Won’t Die: Make…It…Stop!
Now it’s your turn to suffer with this.
CEMETARY POLKA
Tom Waits
Uncle Vernon, Uncle Vernon
Independent as a hog on ice
He’s a big-shot down there at the slaughterhouse
Plays accordion for Mister Weiss
Uncle Biltmore and Uncle William
Made a million during World War II
But they’re tightwads and they’re cheapskates
And they’ll never give a dime to you
Auntie Mame has gone insane
She lives in the doorway of an old hotel
And the radio is playing opera
All she ever says is go to Hell
Uncle Violet flew as pilot
And there ain’t no pretty girls in France
Now he runs a tiny little bookie joint
They say he never keeps it in his pants
Uncle Bill will never leave a will
And the tumour is as big as an egg
Has a mistress, she’s Puerto Rican
And I heard she has a wooden leg
Uncle Phil can’t live without his pills
He has emphysema and he’s almost blind
And we must find out where the money is
Get it now before he loses his mind
Uncle Vernon, Uncle Vernon
Independent as a hog on ice
He’s a big-shot down there at the slaughterhouse
He plays accordion for Mr. Weiss
Buy Rain Dogs. End my suffering.
Upgraded Blog Code
Ok, the site is now running b2evolution 0.9.1 — took some swearing and a weird hack to get it to work, but we’re back!
If you see any weirdness, let me know.
GrabPERF: Agent Location Disabled
This morning, I asked the ERTW.com measurement location to turn down, as we have completed testing the remote measurement code.
This will have some effect on results going forward, mostly positive. The ERTW.com location had an unusual DNS configuration which was affecting the overall measurement statistics.
I am still recruiting for measurement locations on the West Coast. Drop me an e-mail or leave a comment if you are interested in hosting a measurement location on your linux server.
Fatal Flood: Mississippi River Flood of 1927
I watched part of Fatal Flood last night. It was brutal. And, unfortunately, was echoed far too recently not to be a disturbing insight into the nature of the American mind.
The attitude in this nation towards the disadvantaged and disenfranchised has not changed since 1927. This country is falling away from the city upon a hill, the emblematic vision that fuels the misguided and selfish leadership of the United States.
Katrina exposed the raw, festering wounds in the United States. The stench from the Delta is the rotting of the American Soul.
GrabPERF: Bob Wyman Riffs About Online Consistency
Bob Wyman of PubSub has a great post on the value of online consistency to companies and services. [here]
This is a critical component that is too often overlooked. The need for speed is an addictive goal. But as I have said before, fast is not enough.
Go, Bob, Go!
GrabPERF: Why EXPLAIN is such a useful SQL tool
Now, EXPLAIN seems like such an old school SQL tool. Well, I am here to explain why it’s not something you should ignore.
The Index Chart queries were all doing table scans. This can be pretty painful; ok, this can be stupid and EXTREMELY painful. The cause: I was setting sub-optimal date ranges on my queries and making all kinds of crazy date_format calls to format the dates IN THE QUERIES.
Once I changed the code, even when the query cache is cleared, the Index Charts now all run at less than 1 second.
Colour me stupid.
GrabPERF: Whole bunch of bug/performance fixes, part 2
Greg Gershman from BlogDigger found a bug in the Index charts just now. Seems that the chart legend was out of whack with the lines on the chart.
Easy fix. Turns out I was submitting the test ids in a random order. The graph generation component was fine with this, but when I went to build the legend, MySQL ordered the dynamic legend build result in a proper ascending fashion.
Simple fix: when I convert the test ids to an array, I sort the array before proceeding.
Thanks Greg!