On Tue, 13 Jun 2006 14:37:39 +0100
"Brew, CAJ (Chris)" <[log in to unmask]> wrote:
> Hi All,
>
> This came up briefly in the face to face and maybe if we have an
> agreed list we can push we can get something done about them.
>
> To kick things off here's my top three:
>
> 1) Pool Accounts (Passable solution with 5 VOs of 50 users, just
> possible with 20 VOs of 200 Users, it'll be a nightmare with thousands
> of users and groups through VOMS and transient VOs).
> 2) Group Quotas on Storage Elements (though I understand this is
> coming) 3) Logging (Lack of common approach, Large volumes of useless
> info, "Normal" errors in logs and e-mails)
>
> Now's your chance to vent!
>
> Chris.
Not being a system admin my vote counts less and are different from
everyone else.
1) Security that you cant hijack with root.
2) http as a transport protocol.
3) An SRM in front of SRM's to unify name space, and so simplify SRM.
4) All worker nodes behind a NAT with _NO_OUTBOUND_ACCESS_.
5) All major services with full test suites following the model here.
6) Fewer single points of failure.
https://twiki.cern.ch/twiki/bin/view/LCG/LCGgliteTestWritingGuidelines
Regards
Owen Synge
|