Cold Fusion locking up

[PMSLIC]

I have a real problem. My application is a continuing education site for physicians. We use an external web host. They have us running CF 4.5.1 on an NT. We’re their only client therefore they don’t want to upgrade. Actually… they want to discontinue hosting CF by year end. Anyway the powers that be want it rewritten in .net to conform with other projects... but I’m getting off track. There are different databases. One which I use for storing valid license number to verify against. This one gets deleted Monday mornings, and empty one copied and renamed then a CF template ftp’s refreshed data and loads the database. Another stores the course content and physicians scores. Content is read, scores are inserted. No updating of records occur. Scores could be read to see if the physician has already taken the test.

My problem is that from time to time the CF services lock up. There’s no logical pattern. It might go two without happening. Then it might happen in a half hour. Most times when it happens I know they’ve gotten to a certain point because it emails me (this was put in to try to figure out another problem that only happens from time to time) I’ll call it a midway point. When the course is finished it send us another email letting us know that a physician completed it, his score and his evaluation answers. When the services locks/freezes we usually don’t get this last email.

Can anyone offer any suggestions?

 

[NorthStarDA]

when the services lock up, what corrective action is taken to get them responding again? Do you have to restart CF, and have you run a process/memory monitor while doing this to see what is going on?


=========================================
I have not failed. I've just found 10,000 ways that won't work.
Thomas A. Edison

 

[ecobb]

Here's an FAQ on some of the things that can cause CF to lock up. faq232-3662

Let us know if any of these help or if you're still having problems.



Hope This Helps!

Ecobb

"My work is a game, a very serious game." - M.C. Escher

 

[PMSLIC]

NorthStarDA -
The site is hosted externally. When we know the site is locked we email or call the host who restarts the CF Services on the server. They also have a scheduled job which runs a batch job at 2:01 am called 'cycleCf.bat' "which essentially does what they do when they restart the services". (their words)
Also since they are remote I can't monitor the memory/processes.

 

[NorthStarDA]

well i would definitely check out the FAQ to see if maybe some of your code could be causing it, if not- i would get my host to step up and give me some assistance. only CF hosting client or not, it's their responsibility to make sure the server is stable.

i probably would have switched hosts already based on the fact that you are a current customer of CF hosting and it doesn't sound like they are serious in supporting you. I would think they would at least refer you to someone they know as a reputable host if they are not going to support CF anymore.


=========================================
I have not failed. I've just found 10,000 ways that won't work.
Thomas A. Edison

 

[PMSLIC]

Our main site is also hosted there. The communications department manages that and they are using frontpage. They also support some other product they sold us but I'm not involved with that.
Because the powers that be want me to rewrite in .net they aren't going to move.
The support they give does suck. Because we're still on the NT server even the frontpage extensions are at 98 level. We asked if they could update them and they said it would be billable.
A while back they emailed me and told me they were having a problem with emails going out after they did somekind of upgrade on their end. Now other than messing with my local single user development on my own pc I had to tell them how to change the settings on the server.
Anyway... back to the issue at hand. I looked over the faq's. Two items that may contribute is that we do email from the site and we do a fairly large update monday mornings at 6am. I FTP four files, then delete a database that stores a lot of data that is only read thru the week, copy an empty one and then read the records into it. The resulting database is 23meg. I do it this way because I couldnt find a way to compress the DB after just deleting the records and reloading. But this occurs every monday morning. One small file updates daily. The services are recycled daily at 2:01am so I don't think it has anything to do with that.
There was another glitch we are having. Ocassionally I lose a session variable. The same variable. Here again... this doesnt happen that often. It could go for weeks before it happens again. the variable holds the value of the quizID. But between two templates it sets to 0. As a work around I store the variable in a cookie and when the next page comes up if it's not equal to the cookie I reset it. It doesnt happen near as often as the site locking up.