Sign in to follow this  
Followers 0
thelerner

Quantifying the Database/Constant Errors

By thelerner, in Forum and Tech Support

Recommended Posts

thelerner   

thelerner
Posted (edited)

Wonder if we can find out a pattern if we have one thread about when the bug hits and which system.  I'm suggesting people use this thread to log (after the fact ofcourse) when the system was down and what computer and software they use.

 

ie. sample- Got bad Database message twice yesterday, 6/5  6pm & 7pm central time, then gave up.  My system is mac mini running Firefox.

 

If we get enough data in one place, maybe we could start finding solutions.  Maybe get a handle on what is causing it.  No chit chatter on this thread, just a log of when the board has gone down and the computer system used logging in. 

Edited by thelerner
  • Like 1

Share this post

Link to post
Share on other sites

Jetsun   

Jetsun
Posted

For me on my android phone at least 50% of all pages I load come up as an error, but I have just worked out that if I click reload in the address bar then the page will usually load

Share this post

Link to post
Share on other sites

Eques Peregrinus   

Eques Peregrinus
Posted

Aside from the SQL errors, there are also caching problems with the page used to display the error. The browser do not even bother to make a request to the server, but just take the page in its cache. If correctly implemented, this mechanism avoid unnecessary traffic for the server and reduce the delay to display a page for the client.

 

A consequence is that if one time a page is not displayable due to an SQL error, later, even when the server is completely available and the page could be rendered without problems, the browser will fetch the error page in its cache and display it.

 

It is possible to observe if there is a cache problem either by observing the tame taken to display the page by the browser or by sniffing the packets in our private network (only allowed if you own the network).

 

A work around is either clearing the browser cache (radical solution), or by requesting the browser to reload the page without checking its cache (CTRL + F5 for most browsers).

  • Like 4

Share this post

Link to post
Share on other sites

thelerner   

thelerner
Posted

A work around is either clearing the browser cache (radical solution), or by requesting the browser to reload the page without checking its cache (CTRL + F5 for most browsers).

Thanks.  It took me a few steps and many clicks to clear my browser.  If this works universally it'll be a great time saver.

Share this post

Link to post
Share on other sites

Eques Peregrinus   

Eques Peregrinus
Posted

There was a SQL error message yesterday. I took the opportunity to listen to the requests made with the browser and what the server was answering.

 

This is the server's response.

HTTP/1.1 503 Service Unavailable
Server: nginx
Date: Fri, 10 Jun 2016 12:35:35 GMT
Content-Type: text/html
Content-Length: 3049
Connection: keep-alive
X-Powered-By: PHP/5.5.9-1ubuntu4.17
Cache-Control: max-age=864000
Expires: Mon, 20 Jun 2016 12:35:35 GMT

<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
...

We observe the server tells the browser to put the page in cache for 10 days (86400 seconds). For that time, the browser will consider the error page up-to-date, even if the server is working correctly. This is why it is necessary to either to clear the browser cache or to bypass it while reloading the page. (An alternative would be waiting 10 days.)

  • Like 4

Share this post

Link to post
Share on other sites

Apech   

Apech
Posted

I had this problem all yesterday - Mac Mini using Vivaldi - it wasn't a cache problem because I tried Safari and also checked if web site was down using http://www.downforeveryoneorjustme.com .  Is it possible we are experiencing Denial of Service attacks? or is it a software fault because the site needs upgrading.

  • Like 1

Share this post

Link to post
Share on other sites

Marblehead   

Marblehead
Posted

 Is it possible we are experiencing Denial of Service attacks? or is it a software fault because the site needs upgrading.

Your questions are valid.  Perhaps if answers are found corrective action can be taken.

  • Like 1

Share this post

Link to post
Share on other sites

Stosh   

Stosh
Posted

Ive had the sql issues on both my android and PC, though  I didnt clear anything or wait ten days ,but now  I can get back to certain things now,some issues are still popping up. How can the thing think the time has passed and yet still think that it hasnt sometimes. ( Im no computer wizard obviously:) )

Share this post

Link to post
Share on other sites

C T   

C T
Posted

Ive had the sql issues on both my android and PC, though  I didnt clear anything or wait ten days ,but now  I can get back to certain things now,some issues are still popping up. How can the thing think the time has passed and yet still think that it hasnt sometimes. ( Im no computer wizard obviously:) )

As mentioned, to resume complete access you will need to clear your browsing history, aka cache. 

  • Like 1

Share this post

Link to post
Share on other sites

C T   

C T
Posted

Btw it was down all day here as well. 

 

I wonder why hardly anyone highlights this issue as it occurs to the site's FB page which im guessing will alert Sean immediately.

Share this post

Link to post
Share on other sites

Marblehead   

Marblehead
Posted

As mentioned, to resume complete access you will need to clear your browsing history, aka cache. 

I do that nearly every time I shut down my computer but yesterday the system went totally lost in space.  But someone did fix it.  Likely Sean.

  • Like 1

Share this post

Link to post
Share on other sites

thelerner   

thelerner
Posted (edited)

as per others, yesterday 6/10 down all morning.  I'd cleared my caches a few days earlier.   Clearing them again didn't bring the site back.  Firefox and Safari on Mac.

 

 

P.S.  My Firefox was acting funny (not sending me to error pages) on the Amazon site a few days ago. 

Edited by thelerner

Share this post

Link to post
Share on other sites

Nungali   

Nungali
Posted

I didnt bother coming on after the SQL notice the other day. I didnt want to clear my browser.  Today I was resigned to .... but without doing it I now have access - thats a first !

Share this post

Link to post
Share on other sites

Stosh   

Stosh
Posted (edited)

Well, Ill let you know in a few days whether the cache clearing was required, I still havent done that either.

Edited by Stosh

Share this post

Link to post
Share on other sites

dawei   

dawei
Posted

I figure... If I can't get on the site then the staff can get on the site... and there is nothing to moderate... so a useful staff day off :)

 

  • Like 2

Share this post

Link to post
Share on other sites

Michael Sternbach   

Michael Sternbach
Posted

 

I figure... If I can't get on the site then the staff can get on the site... and there is nothing to moderate... so a useful staff day off :)

 

LOL Well, I don't log in only for seeing if there is mod work to do.

 

I totally agree with those who find that the frequent malfunctions are a bummer. Staff is discussing measures to solve this problem.

Share this post

Link to post
Share on other sites

Michael Sternbach   

Michael Sternbach
Posted
I had this problem all yesterday - Mac Mini using Vivaldi - it wasn't a cache problem because I tried Safari and also checked if web site was down using http://www.downforeveryoneorjustme.com''>http://www.downforeveryoneorjustme.com'>http://www.downforeveryoneorjustme.com .  Is it possible we are experiencing Denial of Service attacks? or is it a software fault because the site needs upgrading.

 

It seems to be the latter.

Share this post

Link to post
Share on other sites
Sign in to follow this  
Followers 0
Go To Topic Listing Forum and Tech Support