[Sugar-devel] Datastore index corruption!

Gary Martin garycmartin at googlemail.com
Thu Jul 29 12:45:25 EDT 2010


On 29 Jul 2010, at 15:10, Bernie Innocenti <bernie at codewiz.org> wrote:

> On Thu, 2010-07-29 at 19:15 +1000, James Cameron wrote:
> 
>> Does this block other activity, or can it be done without impact to user
>> operations on the journal?
> 
> The journal icon does not show up in the frame until the indexing is
> finished and no journal operations are possible. So, I guess, you also
> cannot open activities.
> 
>> This 30-60 seconds; how much of that is CPU time, and how much is I/O
>> wait?
> 
> Don't know, I'd have to measure with a full journal...
> 
> Even without testing I have a very strong feeling that automatic
> reindexing on every boot won't be a feasible path. Perhaps we could
> perform full reindexing every N boots, like fsck on old Linux systems?
> Naa... users would be annoyed by the useless occasional delay and would
> demand a button to fix their journal when they know that some data has
> been lost.

OK... Here's a horrible kludge, but I think it's less of a horrible kludge than placing a big 'click this if you think random s##t has happened' button in one of our primary UIs.     

Make a new control panel for such admin/maintenance actions, FWIW to set the scene, in other OSs I'd imagine this as part of some sort of manual disk fix, validate, repair, check, backup, tool. So let's say we have a new cog icon CP module, there is now plenty of space in there to add some sane explanatory text, and a clear text button. It would set a flag, and then reboot the the machine with the usual dialogue as per some other CP modules, rebuild of index would then be done at boot time (so you don't cause even more of a problem as random open activities die and possibly loose users work). Deployments can then choose to include this CP module in their build or not, and I'd guess it stands a better chance of getting upstreamed.   

I'd still much rather see the effort go into addressing the actual underlying issue, but you did ask for suggestions ;)

Regards,
--Gary

> -- 
>   // Bernie Innocenti - http://codewiz.org/
> \X/  Sugar Labs       - http://sugarlabs.org/
> 


More information about the Sugar-devel mailing list