[Klug-general] kentlug.org down?

J D Freeman klug at quixotic.org.uk
Mon Dec 14 12:49:53 UTC 2009


-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Mon, Dec 14, 2009 at 12:44:17PM +0000, dan attwood wrote:
> 
> > Yes, but comparing a single 128 meg vm on a machine with a 128 meg vm that
> > is sharing with 10 others, is not comparing like with like.
> 
> 
>   But are your machines not set up so that they have a minimum amount of 
> ram each - ie 128meg? So 10 x 128 means the host machine will have  
> about 1.3 gig of ram
>   plus ram for the host. Or are you thin provisioning everything?

Each vm gets a specific amount of ram, it can not go over that amount,
and it will not get less than that amount. Some virtualisation platforms
allow for birstable ram use, whic just leads to badness. 

> Oh, and you may want to see whats going on, the console is showing:
> 
> apache2[12529]: segfault at 00007fffff696ff8 rip 00002aaab1abc185 rsp
> 00007fffff697000 error 6
> apache2[22428]: segfault at 00007fffff696ff8 rip 00002aaab1abc185 rsp
> 00007fffff697000 error 6
> apache2[30904]: segfault at 00007fffff696ff8 rip 00002aaab1abc185 rsp
> 00007fffff697000 error 6
> 
> that's exactly why I contacted tech support - to see if there was something else going on with the machine that I hadn't picked up on and indeed it seems there is

I am not sure what apache is doing, there and the console isn't giving a
time stamp for when that error occured.

The vm is now back up with 256 meg of ram. You may want to rejig the
apache conf for the larger ram.

Related question: how big is the database?

Thanks

Julia
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (GNU/Linux)

iD8DBQFLJjRw42M0lILkmGIRApqNAKDmTirEQWxMriqMAyBie8J5UDh5hgCgtC4B
gr7RBZvb3wGufEHVpZVDQNk=
=9e3/
-----END PGP SIGNATURE-----



More information about the Kent mailing list