Re: More ext3 fileserver woes ...

[Neil Brown <neilb cse unsw edu au>]

On Wednesday June 5, akpm zip com au wrote:
> Neil Brown wrote:
> > 
> > ...
> > But with ext3 I see different numbers.
> > 
> > Jun  6 12:04:44 elfman kernel: buffer on 5 has age 9511
> > Jun  6 12:04:44 elfman kernel: buffer on 5 has age 9512
> > Jun  6 12:04:44 elfman kernel: buffer on 5 has age 9513
> 
> ugh.
> 
> I suppose this will fix?

Well... it certainly changes the symptom nicely:
Jun  6 15:05:07 elfman kernel: buffer on 5 has age 498
Jun  6 15:05:14 elfman kernel: buffer on 5 has age 492
Jun  6 15:05:14 elfman kernel: buffer on 0 has age 420
Jun  6 15:05:19 elfman kernel: buffer on 5 has age 496
Jun  6 15:05:25 elfman kernel: buffer on 5 has age 499
Jun  6 15:05:30 elfman kernel: buffer on 5 has age 493

which is suggestive that it fixes the problem....
But how sure are you that it doesn't introduce another problem????
It's not that I think it will (it looks quite believable to me), but
I'm feeling rather fragile about my fileservers at the moment....

I'll probably try it out for real on Wednesday next week.

Thanks a lot,
NeilBrown
  - from the Land of the Long Weekend.

> 
> --- 2.4.19-pre10/fs/buffer.c~ext3-flushtime	Wed Jun  5 21:39:14 2002
> +++ 2.4.19-pre10-akpm/fs/buffer.c	Wed Jun  5 21:39:22 2002
> @@ -1067,6 +1067,8 @@ static void __refile_buffer(struct buffe
>  		bh->b_list = dispose;
>  		if (dispose == BUF_CLEAN)
>  			remove_inode_queue(bh);
> +		if (dispose == BUF_DIRTY)
> +			set_buffer_flushtime(bh);
>  		__insert_into_lru_list(bh, dispose);
>  	}
>  }
> --- 2.4.19-pre10/fs/jbd/transaction.c~ext3-flushtime	Wed Jun  5 21:39:18 2002
> +++ 2.4.19-pre10-akpm/fs/jbd/transaction.c	Wed Jun  5 21:39:22 2002
> @@ -1101,7 +1101,6 @@ int journal_dirty_metadata (handle_t *ha
>  	
>  	spin_lock(&journal_datalist_lock);
>  	set_bit(BH_JBDDirty, &bh->b_state);
> -	set_buffer_flushtime(bh);
>  
>  	J_ASSERT_JH(jh, jh->b_transaction != NULL);