Hi, I searched for this bug but I didn't found it, so I'll enter it now, even
though I can't reproduce it.
1. Go to the loan detail page.
2. Click View all account activity or View transaction history.
BUG: The initial disbursement is missing.
About 1/10 of the loans (784 of 7543) at Jitegemea have this bug. I haven't
been able to figure out the cause yet. They seem to be evenly distributed in
time, not favoring any particular period.
For these 784 loans, the disbursement date is (properly) recorded in the
loan_account table, but it is NOT in the account_trxn table. I think both links
in step 2 above use the account_trxn table.
Platform: All, OS: All
This is issue 1848 and is fixed in v1.1
Bart, I'm assigning to you. Can you check with Jitegemea that they saw the loan
go from approved to active bad standing without being disbursed?
Adam, you are a investigative genius! Thanks for all the amazing hard work.
I've posed both of your questions to Polly at Jitegemea.
1. Are these reproduction steps consistent with what the Jitegemea
data entry people actually saw happen? They would see loans that
they could not disburse, and only were able to make payments
The scenario in 1. sounds very plausible, but I'm not sure if they recall having
loans that were never actually disbursed (by clicking disburse loan) and yet
were able to go ahead with making payments.
2. Were backdated transactions made for any loans repayments? If so,
then the bug probably applies to most if not all the other 47 loans
also. If we want to we can do data investigation to determine which
loans had the backdated transactions.
Backdated transactions were very common (less so now), so it's probably the cause.
As for a backfill, I've posed the question to Aliya, who can consult GK. I'll
ask Jiteg too. But I certainly can't imagine it would be needed before 1.1.
There was a separate email thread about this. We agreed that no backfill is
necessary at this point-- our deployments have been able to handle this OK
I'm resolving this as "Fixed".
As Adam noted above, this was actually issue 1848 and is fixed in v1.1