Is there a solution to avoiding a manual operation to resume transfers and enjoy all the benefits of a cluster.

Please see the normal behaviour of Interpel. When we give the Intialize & Status of the Interpel, it behaves in the following way.  
When you start the Interpel_v660 with an empty mailbox and non existing supctx.dat,
supctx.dat is created like this:

C:\XFB.Interpel_v660_m217\run_time\data>type supctx.dat
xfer_local_ident = 0
xfer_yday = 355
xfer_sequence = 0
sup_state = 1
chk_state = 1

(After for example a power outage, where Interpel could not have updated supctx.dat, it correctly continues the next transfer with the highest local ID in either supctx.dat, or, if this already exists, xfer.dat +1)

When performing some transfers and closing Interpel correctly, supctx.dat looks like this:

C:\XFB.Interpel_v660_m217\run_time\data>type supctx.dat
xfer_local_ident = 3
xfer_yday = 25
xfer_sequence = 3
sup_state = 0
chk_state = 1

When restarting Interpel, next transfer local ID will be 4.
If you have made two more transfers, so normally, when correctly stopped, the local ID should have been 5 in supctx.dat.

However, when closing the Windows session, Interpel is stopped abnormally, but supctx.dat looks like this:

C:\XFB.Interpel_v660_m217\run_time\data>type supctx.dat
xfer_local_ident = 100
xfer_yday = 25
xfer_sequence = 4
sup_state = 1
chk_state = 1

When restarting Interpel, the next transfer will have local ID 101 and not 6.

Supctx does not jump 100 more, but jumps to the next 100, meaning for example 97 to 100 (only 3 more, and not 100). Besides, if in these particular situations the supctx has been modified (rounded up to the next 100).
It is the normal for the performance reasons.

If you want to purge the all the transfers in the Interpel mailbox, you should intialize it from command prompt.

1. C:\XFB.Interpel_v660_m217\run_time\etc>pelmon init_base -f
2. C:\XFB.Interpel_v660_m217\run_time\etc>pelmon start
3. You can perform a transfer to remote partner, you can see that transfer id is 1.