Re: [support.bytemark.co.uk #702699] FOSM server - out of disk space

[Tim Sheerman-Chase]

Yes please!

I think the pycrocosm software is fairly robust while being something we
can extend in future.

Tim


On 18/06/18 12:54, Ross Scanlon wrote:
> Maybe we need to change over to the new server now.
>
> The hardware is stable and I have not had any other issues with the
> software. Although I would like to make sure I've installed the latest
> version of pgmap and pycrocosm.
>
> I'd like to have another change as well.
>
> New id's for nodes, ways, relations, changesets and users to start at
> a number above the current maximum like we did when first forking from
> osm.
>
> I suggest:
>
> changeset   2000000000
> uid               200000000
> node           2000000000000
> way            2000000000000
> relation      2000000000000
>
> Can this be done simply by changing the appropriate value in db_map
> planet_mod_nextids?
>
> We would also need all fosm.org, except map.fosm.org, changed in dns
> to ip 59.167.172.127
>
> All users would need to migrate to the new server so I'd look at
> sending an email with a link to the migrate page to all current user
> emails (these are in the user database) rather than via the mailing
> list.  Sent as bcc so we don't end up with spaming/virus issues and as
> a one off email.   I'd also look at sending via the mailing list to
> let everyone know that we are changing over to the new server and why.
>
> Cheers
>
> Ross
>
>
>
> On 18/06/18 17:52, Jamie Nguyen via RT wrote:
>> Hi George,
>>
>> We think there's two possibilities at the moment for what's happened.
>> Either:
>>
>> 1. the motherboard or PSU (or some other component of the server) has
>> neared the end of its life and causing the server to be unable to see
>> both of the hard drives.
>>
>> 2. both hard drives have died.
>>
>>
>> We sincerely hope it is the first of those two possibilities. At the
>> moment we are transferring the hard drives to separate hardware to
>> try and identify where the problem lies. We'll keep you updated.
>>
>> I might suggest you start to think your disaster recovery options and
>> get your backups ready to restore from, just in case both hard drives
>> have died.
>>
>>
>> Kind regards,
>>
>