HP3000-L Archives

August 2001, Week 1

HP3000-L@RAVEN.UTC.EDU

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
Gavin Scott <[log in to unmask]>
Reply To:
Gavin Scott <[log in to unmask]>
Date:
Wed, 1 Aug 2001 10:49:36 -0700
Content-Type:
text/plain
Parts/Attachments:
text/plain (55 lines)
Ron writes:
> I'm running Netbase and I keep getting this sync error:
>
> 212 13:08:32 NBP001/Sync problem with data base
> MACORD.MACSDATA.WETSEAL, set MESSAGE-FILE (#38)
> 212 13:08:32 NBP001/WARNING 209: Entry 11 was put in the wrong location
>
> What does this mean??

When shadowing, NetBase performs various integrity checks to look for
evidence that the master and shadow copies of the database may have gotten
out of sync somehow.

When a DBPUT is done to the master copy, NetBase intercepts the call to
dbput, calls the real dbput, then writes the entire record into its
transport stream along with one of the return values from dbput, the record
number that Image selected to hold this record (returned in the Image status
array).

When this record makes it over to the posting process on the shadow machine,
POST will perform the same dbput call that was done on the master system,
and it will then check to see that Image chose the same record number to
hold this record on the shadow system that it did on the master system.

Since the function of Shadowing is to ensure that exactly the same sequence
of calls to Image update Intrinsics is made on each machine, the databases
should be in exactly the same state before each dbput and thus Image should
chose the same record each time to hold the record.

If, however, an extra dbput or dbdelete (or certain kinds of database
reorganization) has been done to one database or the other without having
been captured by NetBase, then the datasets will have a different number of
records or a different free-list which will result in the next dbput that
NetBase *does* see going into the wrong location.

When this happens NetBase alerts you to the fact that things have gotten
screwed up, and it gives up trying to apply updates to this particular
dataset.  Since it doesn't keep track (IIRC) of *which* datasets are out of
sync in its directory (it only knows that some dataset(s) in the database
are out of sync) you can get this or a similar sync error after every time
you restart the posting process as long as the database is out of sync as it
rediscovers the fact that this particular set is out of sync.

Sync problems like this are the most fundamental problem one can encounter
with shadowing, and generally indicate some problem either in the
operational procedures being used, the installation of NetBase, or the
configuration of the directory.  Problems can be caused by inappropriately
using an Adager-like tool against one of the databases, restoring a
database, system aborts, etc.

G.

* To join/leave the list, search archives, change list settings, *
* etc., please visit http://raven.utc.edu/archives/hp3000-l.html *

ATOM RSS1 RSS2