Re: check point segments leakage ?

Bruce Momjian wrote:

> Gaetano Mendola wrote:
>
>>Bruce Momjian wrote:
>>
>>
>>>Scott Marlowe wrote:
>>>
>>>
>>>>>>I use a checkpoint_segments = 16 but in my pg_xlog I have
>>>>>>35 files. Why 35 files ?
>>>
>>>
>>>You have 35 because the max files in pg_xlog is 2*checkpoint_segments +1
>>>or something like that. This is documented in the SGML.
>>
>>Ok, that explain why. And they will remain there also if not needed ?
>
>
> Yes, it keeps them around so it doesn't need to recreate them.
>
>
>>Another weird behaviour is that during the day the storage space usage
>>increase gruadualy. Since today as the graph show the space usage
>>is constant, it's like if some space was pre-allocated and now is
>>used, see same yesterday period between 18:00 and 24:00.
>>Toughts ?
>
>
> My guess is that you need a certain amount of free space in the tables
> to operate properly.

Well, today I stop the pg_autovacuum and I did a vacuum full and I reindexed
all big tables and other 500 MB were reclamed. Could be the pg_autovacuum
running yesterday the responsible for these 500MB not reclamed during
a vacuum full and reindex already performed yesterday ?

I'm wandering if will be possible in the 7.5 start and stop the the
autovacuum integrated in the backend.

I don't know if there is space for improvements but add columns to a table
with milion rows is very painfull, for sure could be usefull to do the
following tree operation in one shot:

1) Add column
2) Update the column
3) Set not null

Regards
Gaetano Mendola