[CentOS] Mirror failure

Wed Aug 14 21:05:01 UTC 2013
m.roth at 5-cent.us <m.roth at 5-cent.us>

Phil Dobbin wrote:
> On 14/08/13 21:57, m.roth at 5-cent.us wrote:
>> Phil Dobbin wrote:
>>> On 14/08/13 21:35, m.roth at 5-cent.us wrote:
>>>> Phil Dobbin wrote:
>>>>> I was getting:
>>>>> `http://mirror01.th.ifl.net/epel/6/x86_64/repodata/c60f7c3ee6f9a4902d5ce9dd181a84ca684bba1a1df1c612702c7c6760a04645-filelists.sqlite.bz2:
>>>>> [Errno 14] PYCURL ERROR 6 - "Couldn't resolve host
>>>>> 'mirror01.th.ifl.net'"
>>>>> Trying other mirror.
>> <snip>
>>>>> from epel: [Errno 256] No more mirrors to try.
>>>>>     You could try using --skip-broken to work around the problem`
>>>>> every time I ran `sudo yum update'. This happened on Fedora 17 also.
>>>>> I've had to switch both machines to Ubuntu because I need working
>>>>> machines.
>>>>> I've Googled this extensively & changed my nameservers, commented out
>>>>> relevant lines in the configs, etc, etc but no luck so far. Ubuntu
>>>>> works fine & I have no network problems (two HP ProCurves 2124s
>>>>> working normally over the rest of the network: 17 machines).
>>>>> Is there a workaround for this as I need a working CentOS box.
>>>> You *did* do yum clean all, correct?
>>> No. Was that all it took? Of the thousand answers found on Google that
>>> didn't work, that wasn't mentioned once.

>> Did you try it?

>>> If you could explain why that would work I'd be eternally grateful &
>>> all ears.

>> Your google fu needs work. It should have found a zillion hits.
>> What happens is that yum caches the mirror addresses, and uses the
>> cached addresses, rather than look them up every bloody time. Clearing
>> including the cache, will force it to look Out There again.
> Not a hit on Google concerning this. Try it.

I just did
yum "no more mirrors"
in google, and got a ton of hits, all with this answer.