Overpass API developpement

[FRANCHOIS Serge <>]

No need to apologise Roland. I’m grateful you’re looking into the issue and 
try to help.
Next Monday I will dig up logs, variables, and the diff.

Enjoy the weekend
Serge


> On 10 Nov 2016, at 22:07, Roland Olbricht <> wrote:
> 
> Hi,
> 
> I'm sorry for all the inconveiences. I hope that we can figure out what has 
> happened. Resolving similar issues has helped to add a lot of safeguards.
> 
>> First time I pay attention to that. That is indeed strange.
>> Here’s how I built the database, as per instructions :
>> 
>> $ nohup $EXEC_DIR/bin/init_osm3s.sh $PLANET_FILE $DB_DIR $EXEC_DIR --meta &
> 
> Could you please send us the values of the variables? The directories are 
> not so much of interest. What matters here are
> - Where has the planet file come from?
> - What diff (minute, hour, daily) has been added in which order?
> 
> The error message does indeed mean that the database is somehow broken, on 
> the level that Zlib is unable to decompress this particular block.
> 
>> I’ve no access to my overpass VM at the moment. Will dig in the logs on
>> Monday.
>> Where are those timestamps coming from anyway?
> 
> The timestamp of osm_base is always the timestamp of the latest applied 
> diff. More precisely: the value comes from the .state.txt that has come 
> together with that diff.
> 
> Could you please have a look at the apply_osc_to_db.log in the $DB_DIR? 
> Most interesting are the last 100 lines or so, in particular to figure out 
> which files have been applied.
> 
> Best regards,
> 
> Roland
>