Now im really in trouble. the site wont stay up for more an a few minutes, the cpu maxes out almost immediatly and management start screaming.
I dont know what any of the log files mean so ill post snippets in the hope someone does.
All i know is the w3wp.exe is using 100% cpu and the urls that triggered it every time (without exception) are nb store urls for products that dont exist.
I thought i had solved this by entering the urls that trigger this in the exclusions of the robots file, and for a day it was stable. the next day it fell over and stayed that way since. My next steps are to backup the db and files and remove nb store. if it still happens its obviously not nb store.
Worker process :
In iis7 all i see are a list of urls for products that dont exist and each one has been running in excess of 213986 (it doesnt say seconds or what units that is) but when i used IIS peek the other day it too confirmed that all the operations were for nb store urls.
windows event log (many of)
"A process serving application pool 'DefaultAppPool' suffered a fatal communication error with the Windows Process Activation Service. The process id was '4080'. The data field contains the error number."
system32\logfiles\httperr file (many of)
2013-03-22 17:08:20 66.249.75.134 50602 88.208.220.50 80 HTTP/1.1 GET /Products/SoilRock/tabid/104/ProdID/276/CatID/7/Consolidation_Cell_50mm.aspx - 4 Connection_Abandoned_By_ReqQueue DefaultAppPool
2013-03-22 17:08:20 66.249.75.134 38326 88.208.220.50 80 HTTP/1.1 GET /Products/SoilRock/tabid/104/ProdID/232/CatID/7/CBR_Adjustable_Plunger_Head.aspx - 4 Connection_Abandoned_By_ReqQueue DefaultAppPool
2013-03-22 17:08:20 66.249.75.134 44437 88.208.220.50 80 HTTP/1.1 GET /Products/SoilRock/tabid/104/ProdID/305/CatID/7/2Channel_Junction_Box_for_PC_Interface.aspx - 4 Connection_Abandoned_By_ReqQueue DefaultAppPool
Scheduler:
I have turned off Search Engine Scheduler completly to see if its that, its not.
DNN Event log
There are zero errors except for the application starting every time i have to end task the process as everything has utterly locked up and iis wont restart cleanly.
Application Pool:
I even set the application pool to recycle every 20 minuts - no effect
Server/NB Store:
.Net 4.0, enable 32 bit = false, pipeline= integrated.
nb store = 2.3.2
server os = windows web server 2008 r2 (64bit), i5 dual core dedicated server with nothing else but a noddy static html site running on anothe ip.
8gb memory (using less than 4, even when cpu 100%)
ive the validated storem recycled app pools and anything else i could think of.
Every single time anyone visits one of the urls of a product that doesnt exist (same ones google are visiting) the cpu use goes up 25% and obviously it doesnt take long for it to max the cpu. This is now a serious problem that has me massivly worries since the site is down and will stay that way until this is resolved.
If i cant get it resolved over this weekend i will have no choice but remove nb store from the site, if i do that and it stays up, the boss is going to shoot me for using it in the first place.
Can someone show me how to turn on all forms of diagnostics, logs, stack traces and anything else that might show the cause of this, anything at all no matter how much of a long shot, ill try it.
Any help, is much appreciated as im in deep dodo at this point.
If there is an admin about i am willing to provide either server and/or host access via a skype pm, theres nothing private on it at this point and this has to be solved one way or another!!!
I dont know what any of the log files mean so ill post snippets in the hope someone does.
All i know is the w3wp.exe is using 100% cpu and the urls that triggered it every time (without exception) are nb store urls for products that dont exist.
I thought i had solved this by entering the urls that trigger this in the exclusions of the robots file, and for a day it was stable. the next day it fell over and stayed that way since. My next steps are to backup the db and files and remove nb store. if it still happens its obviously not nb store.
Worker process :
In iis7 all i see are a list of urls for products that dont exist and each one has been running in excess of 213986 (it doesnt say seconds or what units that is) but when i used IIS peek the other day it too confirmed that all the operations were for nb store urls.
windows event log (many of)
"A process serving application pool 'DefaultAppPool' suffered a fatal communication error with the Windows Process Activation Service. The process id was '4080'. The data field contains the error number."
system32\logfiles\httperr file (many of)
2013-03-22 17:08:20 66.249.75.134 50602 88.208.220.50 80 HTTP/1.1 GET /Products/SoilRock/tabid/104/ProdID/276/CatID/7/Consolidation_Cell_50mm.aspx - 4 Connection_Abandoned_By_ReqQueue DefaultAppPool
2013-03-22 17:08:20 66.249.75.134 38326 88.208.220.50 80 HTTP/1.1 GET /Products/SoilRock/tabid/104/ProdID/232/CatID/7/CBR_Adjustable_Plunger_Head.aspx - 4 Connection_Abandoned_By_ReqQueue DefaultAppPool
2013-03-22 17:08:20 66.249.75.134 44437 88.208.220.50 80 HTTP/1.1 GET /Products/SoilRock/tabid/104/ProdID/305/CatID/7/2Channel_Junction_Box_for_PC_Interface.aspx - 4 Connection_Abandoned_By_ReqQueue DefaultAppPool
Scheduler:
I have turned off Search Engine Scheduler completly to see if its that, its not.
DNN Event log
There are zero errors except for the application starting every time i have to end task the process as everything has utterly locked up and iis wont restart cleanly.
Application Pool:
I even set the application pool to recycle every 20 minuts - no effect
Server/NB Store:
.Net 4.0, enable 32 bit = false, pipeline= integrated.
nb store = 2.3.2
server os = windows web server 2008 r2 (64bit), i5 dual core dedicated server with nothing else but a noddy static html site running on anothe ip.
8gb memory (using less than 4, even when cpu 100%)
ive the validated storem recycled app pools and anything else i could think of.
Every single time anyone visits one of the urls of a product that doesnt exist (same ones google are visiting) the cpu use goes up 25% and obviously it doesnt take long for it to max the cpu. This is now a serious problem that has me massivly worries since the site is down and will stay that way until this is resolved.
If i cant get it resolved over this weekend i will have no choice but remove nb store from the site, if i do that and it stays up, the boss is going to shoot me for using it in the first place.
Can someone show me how to turn on all forms of diagnostics, logs, stack traces and anything else that might show the cause of this, anything at all no matter how much of a long shot, ill try it.
Any help, is much appreciated as im in deep dodo at this point.
If there is an admin about i am willing to provide either server and/or host access via a skype pm, theres nothing private on it at this point and this has to be solved one way or another!!!