Print Engine Error 2

[Frobisher]

We've got a problem with CE not running for very long before it starts to generate errors and won't serve up pages anymore. Crystal displays the error: Crystal Reports: Print Engine Error [On Page Server: <x>.pageserver]

The pageserver log shows the following:

Timestamp (Process ID, Thread ID), (Event Type, Category, Event ID), Information:
10:24:9 (905, 877), (1, 1, 33009),  Info: assert failure (Y:\backend\jobserver\dlls\procreport\src\reportdllerrors.cpp699).  (0 : PrintEngineErrorCode: 685).  
Call Stack Dump: Not available

If I bounce the pageserver it will start running again for a while but then stops working soon after.

Any ideas?
TIA

 

[pemad]

I have the same problem. Did you find any solution? I am not having any luck finding a solution. I noticed if I bounce the Crystal Page Server everything is okay until it happens again!!!

 

[Frobisher]

I'm still talking to Crystal support about this problem. They seem to think it's related to running out of resources, whatever that means. I'll let you know when/if I get a solution.

Cheers.

 

[pemad]

Thanks, I also contacted Crystal and they could not give me a answer. I was told to turn logging on as follows: In the command line add at the end of it the following: -loglevel=9 , It has not given me any more information on the error. It looks like the Crystal Page Server is going down when this error occurs. I re-start the page server and all is well again. If I find out anything on this error I will let you know too.

 

[Frobisher]

So is that the way you set the logging level? I couldn't find it documented anywhere. It's funny you were told to do that as I was told:-

&quot;Another thing that can cause this type of behaviour is logging. If you are running the Page Server at loglevel 9 or an extended period of time it could use up a lot of resources on the system and generate errors similar to the one you are seeing.&quot;

I've also just been given this list of things to check, you might want to try them and see if they have any effect, let me know how you get on.

1. Make sure your reports Are NOT based on a crystal .qry file or dictionary file.

2. Make sure your only using System DSNs, on the web server. If you are using odbc.

3. Also check your registry to make sure they really are system DSNs. (regedt32.exe, hkey_local_machine, Software, Odbc, odbc.ini, then highlight your dsn and select security and permissions. Make sure the everyone account is listed in there with full control) Make sure you propagate the everyone account, in that entire DSN's key.

4. Double check to make sure your reports work correctly in the designer on the web server. Basically go infront of the web server and open up the report in the designer and run them. Do they run fine without error?

5. Go into your OS's Services and check to make sure your Crystal Page Server is running on a user account with domain Admin rights.

6. Go into your Page Server service in the CMC, and try setting your Job idle time to a smaller number.

7. To set or un-set, the loglevel parameter you need to stop the service in the CCM and check the properties of the service that you want to check.

Cheers.

 

[pemad]

Thanks for the update. I just received an email from Crystal telling me to check the following.

Check the harddrive space available on where the Web component server
stores it temp files.
It could be that there is not enough space available and that it gets full
in a short time.
So in other words, maybe the page server hangs after a hour, because it
can not write any temp pages or has trouble writting cache pages.

I alos suggested maybe setting the temp and the tmp variables within the
system environment of your winnt server.
This will greatly improve the overall performance of Crystal enterprise
aswell.
You can do this by right clicking on &quot;my computer&quot; and choose &quot;properties&quot;
and go to the environment tab.
Make sure that these 2 variables are in the system environment part.
(note: by default they are only set in the user environment variables, not
in the system environment)

I went through your list and setup the temp variables.
I am in the testing process now. I will let you know how I make out.


Enjoy and Thanks!

 

[dcm]

Frobisher,

We've found that other clients have eliminated similar problems through a programme of continuous clearance of cache and temp files - this is why Crystal suggest not to use qry or dc5 based reports because they generate vast amounts of temp files. As each file requires a handle and will occupy disk space these can have a serious effect on both RAM and space - try monitoring both with performance monitor dor example and see hwat the Page Server is doing.

Also see if you can see whether redundant EPF files are being retained - these should be deleted after a short period of inactivity (I think it's about 20 minutes) when the session closes. David C. Monks
david.monks@chase-international.com

http://www.chase-international.com

Accredited Crystal Decisions Enterprise Partner

 

[pemad]

Crystal Support told me to do the following and it seems to be working. No more 685 error yet!!!!!!


The 685 error can mean a lot of things; but it is an error that has
something to do with troubles in the generating of the report.
You are mentioning idle time settings:
I was recently notified that there is an issue at the moment with the idle
time setting for the page and the cache server.
CE is actually seeing it as seconds and not as minutes.
Try changing the idle time setting for both the cache and the page server
to 10000 instead.
It has been known to solve several issues, hopefully yours aswell.


Try and see if this takes care of your problem also.

Enjoy!

 

[Frobisher]

Just to tie up this thread a bit...we've given up on trying to get CE to run on that machine so we stuck it on a new machine with a fresh NT install and so far it's been ok.

Draw your own conclusions from that!

Frobisher