Republishing...easier way?

[eyeswideclosed]

BOE 11 with CR 11 reports, managed.
We make frequent changes to reports, so I have to:

-delete the report out of the management system (logon, navigate, step-through, etc, logoff...)
-copy the new version of the report to the working folder
-republish it (logon, navigate, step-through, etc, logoff...)
all of this twice, on 2 mirrored machines, usually at least once a day, sometimes more.

Now, this is not intolerable, but I can't help but wonder if there's an easier way to get a new version of a report into the management system.

 

[synapsevampire]

Form within Crystal Reports, select Save As and in the lower left hand corner you'll see Enterprise, click that and sign in, then overwrite the existing report.

This will preserve historical instances as well.

You'll still need to reset the database connectivity and possibly the parameters through the CMC though.

-k

 

[hilfy]

I'm working on a way to do this programmatically using the API so that I can update multiple reports in one call without having to log in to the CMC and do it manually. It will probably be the end of the week before I have it working, at which point I can post information about how I'm doing it.

-Dell

A computer only does what you actually told it to do - not what you thought you told it to do.

 

[malcolmt]

It might be worth having a look at some 3rd party tools such as Object Manger from APOS Systems. I have been using this with success. It enables you to "promote" new versions and preserve existing database connections, parameter settings, schedules etc.

Malcolm.

 

[eyeswideclosed]

Thanks to all, I guess there's no easy answer. Synapse, we have problems connecting to our BOE machines remotely with CR, some programmers can, some can't, plus we have 2 independent but identical BOE installations, so everything has to be done twice. If the publishing wizard let you republish a report without naming it report(2) that would be some help, but it doesn't, you have to delete it first.

 

[Turkbear]

Hi,
Is there some reason you cannot cluster the 2 BOE systems so you would not have to dupliacte things?





To Paraphrase:"The Help you get is proportional to the Help you give.."

 

[eyeswideclosed]

Management wants them completely independent, so if one server blows up the other one can keep delivering reports without interuption, a pain for me but I see their point.

 

[Turkbear]

Hi,
Thats why we cluster..One goes down, the other still delivers..Only of BOTH went south at the same time would the users see any problems..





To Paraphrase:"The Help you get is proportional to the Help you give.."

 

[eyeswideclosed]

Oh, I'm not too familiar with BOE clustering, I mean we have the servers themselves on a load balancer, but the 2 BOEs are separate complete installations on the 2 boxes. Does the clustering you're talking about share anything, like the SQL database?

 

[Turkbear]

Hi,
Yes..In fact it MUST use the same database..

Basically 2 complete installations that share:
CMS database
FRS Input and Output locations (the first set started takes control with the second used if the first fails)

The BOE system handles both load balancing and fail-over .

All servers (the 2 BOE and the CMS database one) must be on the same subnet to avoid latency and router issues).

The FRS must reside on a file system location that the pageserver(s), job server(s) and cache servers can access.


two papers that detail this type of configuration( refers to CE8.5 but applies to BOE XI as well:

http://support.businessobjects.com/communityCS/TechnicalPapers/ce85_high_availability.pdf.asp

and

http://support.businessobjects.com/communityCS/TechnicalPapers/ce85_multiple_frs.pdf.asp

To Paraphrase:"The Help you get is proportional to the Help you give.."

 

[eyeswideclosed]

I see, I think, thanks. But if the shared cms database exists only on 1 machine and that machine goes down...

 

[Turkbear]

Hi,
That is why we use an Oracle database in a very stable configuration in our Network center - it is running in archivelog mode, so data recovery is not difficult, if ever needed .

We also export the data nightly and do cold backups every week.

We could, at least in theory, be back up with very minimal loss of CMS data in about 1 hour .








To Paraphrase:"The Help you get is proportional to the Help you give.."

 

[eyeswideclosed]

Ok I'm clear now, thanks. They want 2 totally independent yet mirrored systems though, so I am relegated to doing everything twice, ah well, things could be worse I guess.

 

[Turkbear]

Hi,
Yep..Just wait, they will be -

Remember, that which does not kill you only delays the inevitable....( from a great set of demotivation posters :

http://www.despair.com/index.html)

To Paraphrase:"The Help you get is proportional to the Help you give.."

 

[synapsevampire]

Synapse, we have problems connecting to our BOE machines remotely with CR, some programmers can, some can't"

I used to occasionally have difficulties connecting in CE 8.5, but I've never had a problem since.

I suggest that you resolve this problem first, it certainly is a warning of some sort, then you might later address the concern by writing custom code to write to both.

-k

 

[eyeswideclosed]

Yeah, it's network access/permission problems, but even if we got it to work we would still have to modify the report in question on both boe systems, but this would definitely be easier than what I'm doing now, thanks.