damn 30 seconds
06-01-2006, 04:51 PM
#4
But in seriousness, why is it setup that way? Would too many searches slow down the website? Would this site be sped up if people would regularly purge all their sent and received PMs?
06-01-2006, 08:35 PM
#6
Administrator
I'm the first to admit that I don't know **** about this stuff, but I'll take a stab at it.
I believe there are robot-type programs (bots?) that can attack and crash sites
by searching like crazy. So, the 30 second rule stops that.
Help a brother out here Brains.
I believe there are robot-type programs (bots?) that can attack and crash sites
by searching like crazy. So, the 30 second rule stops that.
Help a brother out here Brains.
06-01-2006, 08:44 PM
#7
Originally Posted by RrCoX22
normally it's not 30 secs that you have to wait, but more like 5-10 secs
Trending Topics
06-01-2006, 08:47 PM
#8
SSU's Drunk Typing Whore
iTrader: (47)
Join Date: Jan 2004
Posts: 4,609
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by unit213
I'm the first to admit that I don't know **** about this stuff, but I'll take a stab at it.
I believe there are robot-type programs (bots?) that can attack and crash sites
by searching like crazy. So, the 30 second rule stops that.
Help a brother out here Brains.
I believe there are robot-type programs (bots?) that can attack and crash sites
by searching like crazy. So, the 30 second rule stops that.
Help a brother out here Brains.
06-01-2006, 08:52 PM
#9
TECH Enthusiast
Thread Starter
iTrader: (1)
Join Date: Dec 2005
Location: Farmington NH
Posts: 597
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by unit213
I'm the first to admit that I don't know **** about this stuff, but I'll take a stab at it.
I believe there are robot-type programs (bots?) that can attack and crash sites
by searching like crazy. So, the 30 second rule stops that.
Help a brother out here Brains.
I believe there are robot-type programs (bots?) that can attack and crash sites
by searching like crazy. So, the 30 second rule stops that.
Help a brother out here Brains.
its time to hunt down every last one of these damn robots and kill the lot of em
viva la resistance!
06-01-2006, 11:17 PM
#12
TECH Veteran
Originally Posted by SmokeySS
Would too many searches slow down the website?
That's why they turned off search for people who aren't registered / logged in, about six months ago.
Along with a number of other features that were disabled/limited around then, it's done as a temporary workaround until whenever things really get fixed.
06-02-2006, 08:25 AM
#13
TECH Senior Member
iTrader: (7)
Join Date: Jan 2002
Location: Katy, TX
Posts: 12,754
Likes: 0
Received 0 Likes
on
0 Posts
Its hard to really relate the complexities of running a site this large to some folks, until they're actually the ones doing the work. A lot of the things that seem pretty simple, end up being the biggest headaches. Search is one of those things that seems trivial, but ends up being huge.
Think of it this way -- you are a file clerk at a company with 60,000 employees, and over 4 million documents in a filing center. Throughout the course of your work day, most of these employees request documents from you one at a time, and request them by a unique ID number. You find the document and hand it to them. Now, you're also tasked with finding documents for folks by what's written on them. Joe from Accounting says "I need all documents that talk about 'big blue widgets'." So, you go off to look through the documents. In the mean time, all the regular employees are still asking you for documents. Now Mary from Purchasing wants all the documents written by Fred in Logistics, and then Buford from Sanitation wants every document about trash cans. All the while, you're still expected to get everyone their documents by number.
This is exactly what we're up against with search. Finding things by number (ie. post id or thread id, the number displayed in the address bar) is very fast and easy. Finding things by the words INSIDE that document is not -- the system has to read through the documents (rather, indexes of the documents) to find what you want, make a list of ID numbers, and then return the results. All the while, its still handing out regular documents.
Some folks have even said "well Google does it, why can't you?" We have 4 servers now. Google has, at the last estimate over 2 years ago, well over 100,000 servers. Its a simple matter of scale.
Think of it this way -- you are a file clerk at a company with 60,000 employees, and over 4 million documents in a filing center. Throughout the course of your work day, most of these employees request documents from you one at a time, and request them by a unique ID number. You find the document and hand it to them. Now, you're also tasked with finding documents for folks by what's written on them. Joe from Accounting says "I need all documents that talk about 'big blue widgets'." So, you go off to look through the documents. In the mean time, all the regular employees are still asking you for documents. Now Mary from Purchasing wants all the documents written by Fred in Logistics, and then Buford from Sanitation wants every document about trash cans. All the while, you're still expected to get everyone their documents by number.
This is exactly what we're up against with search. Finding things by number (ie. post id or thread id, the number displayed in the address bar) is very fast and easy. Finding things by the words INSIDE that document is not -- the system has to read through the documents (rather, indexes of the documents) to find what you want, make a list of ID numbers, and then return the results. All the while, its still handing out regular documents.
Some folks have even said "well Google does it, why can't you?" We have 4 servers now. Google has, at the last estimate over 2 years ago, well over 100,000 servers. Its a simple matter of scale.
06-02-2006, 09:14 AM
#14
TECH Resident
iTrader: (48)
Join Date: Nov 2001
Location: Michigan, the land of cold
Posts: 927
Received 0 Likes
on
0 Posts
Google is also a company (publicly traded no less), not an enthusiast board. I'd believe that their budget is slightly larger than the boards.
You can use Google to search this board. I've done it before.
Put "STS turbo" site:ls1tech.com into Google, it searches THEIR database. It's fast and it doesn't hurt the board. Granted it will not give you the newest stuff, but if you are searching for many instances of a given topic it works well.
You can use Google to search this board. I've done it before.
Put "STS turbo" site:ls1tech.com into Google, it searches THEIR database. It's fast and it doesn't hurt the board. Granted it will not give you the newest stuff, but if you are searching for many instances of a given topic it works well.
06-02-2006, 09:36 PM
#16
TECH Veteran
Originally Posted by SmokeySS
So what about people regularly clearing out PMs?
Search is the major problem.
PMs aren't searched.
So no. Don't worry about it.
06-02-2006, 09:37 PM
#17
TECH Veteran
Originally Posted by Brains
Its hard to really relate the complexities of running a site this large to some folks, until they're actually the ones doing the work. A lot of the things that seem pretty simple, end up being the biggest headaches. Search is one of those things that seems trivial, but ends up being huge.
Think of it this way -- you are a file clerk at a company with 60,000 employees, and over 4 million documents in a filing center. Throughout the course of your work day, most of these employees request documents from you one at a time, and request them by a unique ID number. You find the document and hand it to them. Now, you're also tasked with finding documents for folks by what's written on them. Joe from Accounting says "I need all documents that talk about 'big blue widgets'." So, you go off to look through the documents. In the mean time, all the regular employees are still asking you for documents. Now Mary from Purchasing wants all the documents written by Fred in Logistics, and then Buford from Sanitation wants every document about trash cans. All the while, you're still expected to get everyone their documents by number.
This is exactly what we're up against with search. Finding things by number (ie. post id or thread id, the number displayed in the address bar) is very fast and easy. Finding things by the words INSIDE that document is not -- the system has to read through the documents (rather, indexes of the documents) to find what you want, make a list of ID numbers, and then return the results. All the while, its still handing out regular documents.
Some folks have even said "well Google does it, why can't you?" We have 4 servers now. Google has, at the last estimate over 2 years ago, well over 100,000 servers. Its a simple matter of scale.
Think of it this way -- you are a file clerk at a company with 60,000 employees, and over 4 million documents in a filing center. Throughout the course of your work day, most of these employees request documents from you one at a time, and request them by a unique ID number. You find the document and hand it to them. Now, you're also tasked with finding documents for folks by what's written on them. Joe from Accounting says "I need all documents that talk about 'big blue widgets'." So, you go off to look through the documents. In the mean time, all the regular employees are still asking you for documents. Now Mary from Purchasing wants all the documents written by Fred in Logistics, and then Buford from Sanitation wants every document about trash cans. All the while, you're still expected to get everyone their documents by number.
This is exactly what we're up against with search. Finding things by number (ie. post id or thread id, the number displayed in the address bar) is very fast and easy. Finding things by the words INSIDE that document is not -- the system has to read through the documents (rather, indexes of the documents) to find what you want, make a list of ID numbers, and then return the results. All the while, its still handing out regular documents.
Some folks have even said "well Google does it, why can't you?" We have 4 servers now. Google has, at the last estimate over 2 years ago, well over 100,000 servers. Its a simple matter of scale.
