[lug] slocate.db.stf

Ben bluey at iguanaworks.net
Thu Dec 28 08:50:25 MST 2006


>> Everyday for the past week I've been getting this e-mail from
>> cron.daily:
>>
>> /etc/cron.daily/slocate:
>> updatedb: fatal error: The temp file '/var/lib/slocate/slocate.db.stf'
>> already exists and does not appear to be a valid slocate database.
>> Please remove before creating the database.
>>
>>
>> I'm running debian sarge and I swear I didn't change anything. I just
>> started getting the message. There is
>> no /var/lib/slocate/slocate.db.stf on the harddrive now, and when I
>> run /etc/cron.daily/slocate from the command prompt, I don't get any
>> errors. The slocate database is being updated, so maybe I can ignore
>> the message. Any ideas what's going on?
>>     
>
> Could be more than one copy of updatedb trying to run at the same time.
> Check your crontabs and cron logs for duplicates.
>   
Ok, so a duplicate cron entry sounds like reasonable cause. The .stf is 
a temp file that goes away when the slocate cron finishes, which is why 
I don't see it normally, but it exists when this second cron comes up 
and then gives me that error. So how can I test this? All the entried in 
/etc/cron.daily/* run sequentially, right? The only related cron I see 
is find.notslocate. I made it non-executable, but I'm getting the same 
error. I checked /etc/cron.d/* I don't see anything slocate related. My 
logs don't show anything helpful about cron -- just that cron.daily ran. 
And no idea why this would have started all of a sudden. If it were a 
program crash or something, shouldn't I be able to find the stf file. If 
a duplicate entry for cron, maybe now it is taking longer to run and 
bumping up against the second cron and before it ran faster and they 
didn't overlap in time?

Thanks,

Ben





More information about the LUG mailing list