Translating with DrWatson… this can take a few seconds the first time.
This is a genuine, complex translation. DrWatson protects commands, error codes, and log output while naturally translating the surrounding text. It’s translated once and saved.
Re: 7.31 performance problem
Answered: green (solid confidence) — Andrew Hamm questioned whether LBU_PRESERVE only matters for onarchive; David Williams corrects this (they hit it with ontape too) and Doug McAllister explains the mechanism (archiving a log requires logging that fact, so the last log must be preserved); asker accepts and keeps the setting.
Andrew Hamm asked whether the long-transaction-related parameter in 7.31 applies only to onarchive. David Williams said it applies when logical logs fill, whatever the backup tool, and that his team hit it with ontape. Doug McAllister explained the last log is preserved so that archiving the other logs can itself be logged. Hamm kept the setting.
Auto-generated by Claude from the posts below — may be imperfect; read the full thread.
Andrew Hamm — — source: Informix-list mailing list archive (1991-1998)
David Williams wrote in message ...
>
>>LBU_PRESERVE 0 # Preserve last log for log backup
>
> Set to 1!
Emmm, excuse me for going sydwayz here, but isn't this parameter only
relevant if you are foolish enough to use onarchive for backups? I
understand that onarchive writes to sysutils database leading to a potential
log deadlock, but ontape and onbar don't have this problem.
I must confess to having your opinion too, but recently when reading the
fine print, it sounded like it was only relevant for onarchive.
↪ replying to Andrew Hamm
David Williams — — source: Informix-list mailing list archive (1991-1998)
In article <3a5274df$1@news.iprimus.com.au>, Andrew Hamm
<ahamm@sanderson.net.au> writes
>David Williams wrote in message ...
>>
>>>LBU_PRESERVE 0 # Preserve last log for log backup
>>
>> Set to 1!
>
>Emmm, excuse me for going sydwayz here, but isn't this parameter only
>relevant if you are foolish enough to use onarchive for backups? I
>understand that onarchive writes to sysutils database leading to a potential
>log deadlock, but ontape and onbar don't have this problem.
>
>I must confess to having your opinion too, but recently when reading the
>fine print, it sounded like it was only relevant for onarchive.
>
>
It is used when the logical logs are filling up and is not specific
to the tool used to back them up. We have hit this using ontape.
>
--
David Williams
↪ replying to David Williams
Andrew Hamm — — source: Informix-list mailing list archive (1991-1998)
David Williams wrote in message ...
>
> It is used when the logical logs are filling up and is not specific
> to the tool used to back them up. We have hit this using ontape.
>
I started setting it 'cos I was involved in the recovery of a few deadlocked
long transaction problems (don't ask, I'm still crying years later) and I
figured it sounded like a little bit of a safety valve.
So I'll just just keep that fella in there, hey?
Cheers
↪ replying to Andrew Hamm
Doug McAllister — — source: Informix-list mailing list archive (1991-1998)
When you archive a log file to tape, Informix needs to log that fact. If
the logs are already full, you cannot log the log archive. Therefore, by
preserving the last log, the archive of all of the other log files can
be logged.
Sounds like a tongue-twister doesn't it?
:)
Andrew Hamm wrote:
>
> David Williams wrote in message ...
> >
> > It is used when the logical logs are filling up and is not specific
> > to the tool used to back them up. We have hit this using ontape.
> >
> I started setting it 'cos I was involved in the recovery of a few deadlocked
> long transaction problems (don't ask, I'm still crying years later) and I
> figured it sounded like a little bit of a safety valve.
>
> So I'll just just keep that fella in there, hey?
>
> Cheers
We use strictly necessary cookies to make this site work. With your
consent we’d also use optional cookies for analytics and marketing. You can accept all,
reject all, or choose. Read our Cookie Policy.