Trolling for Wise Ideas

[Guy Werry]

We have a recurring problem, not an urgent one, that I'd like to see if
anyone has Wise Ideas about.

We have an environment running under HP-Ux 10.2, Powerhouse 7.33.D3,
C-ISAM (unsure of version, but it's the latest version that 7.33.D3 came
with) and FORTRAN (ick!).

We have an Image/3000 emulator built by a contractor: it takes the Image
calls and translates them into C-ISAM calls - we've had this since 1994
and it has been very reliable.

This particular application has one large data file (2 million plus
rows) which is heavily indexed: data file currently is about 650 K but
the index is twice that size.

We have a FORTRAN program that does batch updates to the data file,
locking mode 1, around the entire transaction.  This program only adds
to the data file.
We also have Powerhouse screens that allow users to do maintenance on
the data file, which often means deleting records that were uploaded
erroneously.

The Powehouse screens use the default locking scheme ... there is
nothing explicitly written in the code or dictionary regarding locking.

Periodically we come up with corrupted indexes on the main data file and
have to use the bcheck utility to re-build the indexes.  Sometimes this
happens several times in a couple of months, or like this week, it will
be almost a YEAR between occurrences.

I had long thought that we must have the FORTRAN program and the
maintenance screens tripping over each other.  This past week, when the
problem showed up I actually saw 2 users in the maintenance screens as
well as one of the FORTRAN uploads taking place: one of the maintenance
users actually received a message that the screen was waiting on locks.
Then, the corrupted index showed up.

The FORTRAN does lock, although it is a mode 1 lock, which allows other
concurrent users.  

So, what I'm thinking is that our FORTRAN program's locks are not
keeping the maintenance screens out, the maintenance screens are
deleting entries (requireing notoriously slow index rebuilds in C-ISAM)
and the two processes working against each other are resulting in
messed-up indexes.  Does this make sense?

Thanks,
Guy L. Werry
Senior Systems Analyst
Hudson Bay Mining & Smelting Co., Limited. 


The information in this e-mail and any attachments is confidential and may be subject to legal professional privilege. It is intended solely for the attention and use of the named addressee(s). If you are not the intended recipient, or person responsible for delivering this information to the intended recipient, please notify the sender immediately. Unless you are the intended recipient or his/her representative you are not authorised to, and must not, read, copy, distribute, use or retain this message or any part of it.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.sowder.com/pipermail/powerh-l/attachments/20080415/a0b7eb66/attachment.htm