Initial import of ClearCase repo - shortcut?

Hi,
We are planning to synchronize an existing ClearCase repository with a new RTC repo. The ClearCase repo has many files and VOBs (approx 180k files and 20 VOBs). The ClearCase repo and RTC repo will be synchronized across a WAN.
Through experimentation, we have learned that the initial import of the code can be done effectively in a piecemeal manner (see TN0006 - http://jazz.net/library/techtip/68). However, the initial import times are not fast (i.e. it took us 5 hrs to import 14k files - over a LAN). (Note that sync times after initial import are quite speedy - less than a minute for the same number of files :) )
The question we have is if there is a way to effectively "shortcut" the initial import through the synchronizer by simply making of a copy of the ClearCase files and committing them to the RTC repo. Then, we would run synchronization against these files and, in theory, would only every have to communicate changes through the synchronizer.
Is this approach feasible? I am a bit doubtful that this approach would work, but if so it would save us some time and effort. If it is feasible, what constraints should we be aware of (i.e. would files on the ClearCase side have to remain static until the first sync, etc.)?
Thanks,
--Carl
We are planning to synchronize an existing ClearCase repository with a new RTC repo. The ClearCase repo has many files and VOBs (approx 180k files and 20 VOBs). The ClearCase repo and RTC repo will be synchronized across a WAN.
Through experimentation, we have learned that the initial import of the code can be done effectively in a piecemeal manner (see TN0006 - http://jazz.net/library/techtip/68). However, the initial import times are not fast (i.e. it took us 5 hrs to import 14k files - over a LAN). (Note that sync times after initial import are quite speedy - less than a minute for the same number of files :) )
The question we have is if there is a way to effectively "shortcut" the initial import through the synchronizer by simply making of a copy of the ClearCase files and committing them to the RTC repo. Then, we would run synchronization against these files and, in theory, would only every have to communicate changes through the synchronizer.
Is this approach feasible? I am a bit doubtful that this approach would work, but if so it would save us some time and effort. If it is feasible, what constraints should we be aware of (i.e. would files on the ClearCase side have to remain static until the first sync, etc.)?
Thanks,
--Carl
One answer

I have good news and bad news.
The good news is that there is a simple answer to your question.
The bad news is that the answer is "no, that would not help" (:-).
In particular, it actually would be slower to import a tree on top of an
existing tree, because it has to do everything it would do for an
import, but then also compare the content of what it is bringing over
with what is already there (if they are the same, it will not create a
new version).
Cheers,
Geoff
On 8/19/2010 1:07 PM, egghead9 wrote:
The good news is that there is a simple answer to your question.
The bad news is that the answer is "no, that would not help" (:-).
In particular, it actually would be slower to import a tree on top of an
existing tree, because it has to do everything it would do for an
import, but then also compare the content of what it is bringing over
with what is already there (if they are the same, it will not create a
new version).
Cheers,
Geoff
On 8/19/2010 1:07 PM, egghead9 wrote:
Hi,
We are planning to synchronize an existing ClearCase repository with a
new RTC repo. The ClearCase repo has many files and VOBs (approx 180k
files and 20 VOBs). The ClearCase repo and RTC repo will be
synchronized across a WAN.
Through experimentation, we have learned that the initial import of
the code can be done effectively in a piecemeal manner (see TN0006 -
http://jazz.net/library/techtip/68). However, the initial import
times are not fast (i.e. it took us 5 hrs to import 14k files - over
a LAN). (Note that sync times after initial import are quite speedy -
less than a minute for the same number of files :) )
The question we have is if there is a way to effectively
"shortcut" the initial import through the synchronizer by
simply making of a copy of the ClearCase files and committing them to
the RTC repo. Then, we would run synchronization against these files
and, in theory, would only every have to communicate changes through
the synchronizer.
Is this approach feasible? I am a bit doubtful that this approach
would work, but if so it would save us some time and effort. If it is
feasible, what constraints should we be aware of (i.e. would files on
the ClearCase side have to remain static until the first sync,
etc.)?
Thanks,
--Carl