Fixing Projects That Crash

[gmg]

We have a project file in our office that has repeatedly crashed and I am having trouble fixing the problem(s). I did the usual: audit, purge unused, and compact file. I also tried to fix a few unresolved warnings with the review warnings tool. There are still many unresolved issues and it would take a very long time to resolve each warning. That could very well be the problem.

What is everyone else doing to keep your project files stable?

It's worth noting that I rarely have Revit projects crash, so this is unusual.

[greg.mcdowell]

Doesn't sound like your issue but the other day I had a Workset enabled file crash and when I went to the backups they would not open either. I don't ultimately know what was wrong but doing a Save As and selecting Make This a Central File After Save got it working again.

I think you will, unfortunately, have to dig through those errors and start resolving them. Some are certainly more troublesome than others (I think Join errors are particularly suspect).

Good luck!

[gordonp147484]

We have a project file in our office that has repeatedly crashed and I am having trouble fixing the problem(s). I did the usual: audit, purge unused, and compact file. I also tried to fix a few unresolved warnings with the review warnings tool. There are still many unresolved issues and it would take a very long time to resolve each warning. That could very well be the problem.

What is everyone else doing to keep your project files stable?

It's worth noting that I rarely have Revit projects crash, so this is unusual.

We have found that on problem files, it is often good to audit and save to a totally different file. Some people have taken to including the Audit date in the new central file name, and moving the old central to archive, that way everyones local will be broken and they will be forced/reminded to make new locals. Others have been doing a multi-step process to get a clean file with the same name as before. With this approach it is important to make sure everyone makes new locals, as a pre-audit local seemingly can re-corrupt the newly minted central.

Best,
Gordon

[gmg]

Thanks for the suggestion Gordon. I'll give it a try.

Another thing I have wondered about is how many team members on a single project are too many? At one point yesterday, there were size people accessing a single central file. It seems like the more people on a worksharing enabled project, the higher likelihood for errors.

[gordonp147484]

Thanks for the suggestion Gordon. I'll give it a try.

Another thing I have wondered about is how many team members on a single project are too many? At one point yesterday, there were size people accessing a single central file. It seems like the more people on a worksharing enabled project, the higher likelihood for errors.

The more people on a team, the slower everyone is, that is for sure. And I don't know if a larger team creates more chance for a software induced error, but it certainly creates more chance for a user induced error, especially when the team balloons with "whoever is available" to put out fires. Often we throw someone with little Revit experience into this situation, because that is what we did in Acad and hand drafting. It wasn't really a good idea then either, but with Revit it can be catastrophic. One more way that Revit forces people to change their behavior, even when they will never touch Revit, or even a computer. You are making staffing decisions, Revit affects you. You are doing hiring, Revit affects you. You are writing contracts, Revit affects you. Those people are just now beginning to realize this and change accordingly.
One other thing on a technical note. A team of 3 on a 100MB shared file might be ok on a 100MB network connection. But a team of 15 on that same project and connection will be miserable. Gigabit is required at this point, and when even one person is on 100MB I have seen errors, usually for the slow person, but not always.

Gordon

[armbarsalot]

save as a new central file, we have a dozen or so working in a new central file each day, then send it to india where it is re-centralized & worked on by 20+ people. The files stay very healthy, rarely corrupt elements or issues. Very impressed.