Subject | Re: Target shutdown mode is invalid for database |
---|---|
Author | patrick_marten |
Post date | 2013-01-23T16:39:41Z |
Hi Mark,
I did some more research. When the state is "force write" (according to gstat), then starting up the database will fail because of the hierarchy, right?
The articel says "This hierarchy is important, you cannot shutdown a database to a higher or equal level that it currently is, nor can you startup a database to a lower or equal level."
It looks as if the database get's restarted somewhere else already, when I try to restart it, so that the level is "force write" again, which equals to "normal" (I think), i.e. it fails, correct? If yes, then I will have to sort the code of my colleague, I guess and see, where it gets restarted inbetween...
> What doesn't work? Do you get an error message, if so: what is the errorThe log contains the error message in the title (sorry, forgott to mention it in the message too). So "Target shutdown mode is invalid for database"
> message
>
I did some more research. When the state is "force write" (according to gstat), then starting up the database will fail because of the hierarchy, right?
The articel says "This hierarchy is important, you cannot shutdown a database to a higher or equal level that it currently is, nor can you startup a database to a lower or equal level."
It looks as if the database get's restarted somewhere else already, when I try to restart it, so that the level is "force write" again, which equals to "normal" (I think), i.e. it fails, correct? If yes, then I will have to sort the code of my colleague, I guess and see, where it gets restarted inbetween...
> > One version of my application ships with FB 2.1.4, the new one with FBYes, I've re-read it a few minutes after I sent my message and when I read the description of the "normal" state, I figured that I got it wrong about "full".
> > 2.5.2. For the first one it seems to be enough to introduce some delay
> > between the processes, for the second one it doesn't work.
> >
> > The article
> >
> http://www.firebirdsql.org/file/documentation/reference_manuals/user_manuals/html/gfix-dbstartstop.html
> > says that using "full" is not an acceptable option for starting up a
> > database, so I've also tried to shutdown the database using "single"
> > instead, but it didn't change anything.
>
> That document tries to convey that "gfix -shut full" is acceptable, while
> "gfix -online full" isn't, as that doesn't make sense for online.
>
> Mark
>