Can anybody shed light onto what _lock-flag 'I' is?

Posted by neil.fern on 18-Jan-2012 05:05

We have a customer on 9.1e who is getting a lot of _locks with _lock flag I.  Docs only mention Shared, eXclusive, Limbo, Upgrade and Queued. These _lock records have _lock-recid 0 (zero), does this means they have been used but are now Inactive?

Thanks in advance.

All Replies

Posted by asthomas on 18-Jan-2012 05:11

Jeg er væk fra kontoret i dag den 18. januar 2012, og checker ikke altid mail og telefon regelmæssigt. Jeg vil dog svare på din mail så snart som muligt.

Hvis du har behov for support, kontakt venligst, så vender vi tilbage så snart det er mnuligt.

Administrative henvendelser kan ske til :

I am not in the office 18th January 2012. I will not be checking my mail and phone on a regular basis, but will reply to your mail as soon as I can.

If you need support from appSolutions, please send your mails to

If you need Roundtable support, please send mails to :

Med venlig hilsen / Regards

Thomas Hansen

appSolutions a/s

Posted by neil.fern on 18-Jan-2012 05:31

All I think i have found the answer in the thread titled 'Locking with Chain type "TAB" instead of "REC"'

Sorry for the inconvenience.


Neil Fern

Analyst Developer

Sage Construction

Sage (UK) Limited

Posted by ChUIMonster on 18-Jan-2012 08:00

From an old PEG thread (which seems appropriate since this is a 9.1 question):

TAB is indeed a table lock, required for SQL92. There are implicit and
explicit table locks.  You can explicitly request a table lock using SQL92.
An implicit table lock is used when you have a lock on one of the records.

The three types of implicit table locks are:
- Intent-Exlusive (IX) for when you have a exclusive (X) lock on a record.
- Intent-Shared (IS) for when you have a shared (S) lock on a record.
- Shared-intent-exclusive (SIX). This is when you have an EXPLICIT shared
table lock (S) and then implicitly gain an exlusive lock on a record.

Neil Davies
Inforensics Ltd.
Performance Systems and Services

Posted by neil.fern on 18-Jan-2012 08:29

Thanks for the further info Tom.  We now understand the problem better.

This thread is closed