per-entity - Re: [Per-Entity] distribution of aggregate metadata
Subject: Per-Entity Metadata Working Group
List archive
- From: Nicholas Roy <>
- To: IJ Kim <>, Scott Koranda <>
- Cc: "Cantor, Scott" <>, Thomas Scavo <>, Per-Entity Metadata Working Group <>
- Subject: Re: [Per-Entity] distribution of aggregate metadata
- Date: Fri, 12 Aug 2016 15:30:26 -0600
- Authentication-results: spf=none (sender IP is ) ;
- Ironport-phdr: 9a23:jtMTHxCcqH2WN1+T0BvCUyQJP3N1i/DPJgcQr6AfoPdwSP7/pMbcNUDSrc9gkEXOFd2Crakb26yL6Ou5BCQp2tWojjMrSNR0TRgLiMEbzUQLIfWuLgnFFsPsdDEwB89YVVVorDmROElRH9viNRWJ+iXhpRZbIBj0NBJ0K+LpAcaSyp3vj6Hh1oPeeTpIhSawJPZbZFXz9F2J95pevYw3C4R56RrM6i9JecxXw39lP1Seg0y668utqtoryC9Xp+5p1MlaWKP2duxsVr9fFj0gP2kd68jitB2FRgyKsCgySGITxzxJCAjC6BjhFq32o2OusPB6yQGbO9H7V7Y5RW7k4qt2Hky7wBwbPiI0pTmEwvd7i7hW9Vf7/0Ry
- Spamdiagnosticmetadata: NSPM
- Spamdiagnosticoutput: 1:99
On 8/12/16 1:57 PM, IJ Kim wrote:
> On Thu, Aug 11, 2016 at 12:27:19PM -0500, Scott Koranda wrote:
>>>> On Aug 10, 2016, at 12:12 PM, Cantor, Scott
>>>> <>
>>>> wrote:
>>>>
>>>>> I think there would have to be some real benefit (not just experience
>>>>> gained) before we would consider doing that. From where I sit, the
>>>>> current distribution method Just Works (TM) so I have to turn around
>>>>> and ask InCommon deployers: Is there something about the current
>>>>> distribution method that is suboptimal or not working for you (apart
>>>>> from the size of the aggregate of course)?
>>>> I don't think the point was to help us (and nobody is suggesting the URL
>>>> would change), so if Ops doesn't think incorporating a cloud/CDN
>>>> delivery strategy under cover of a name controlled by InCommon is a
>>>> useful step for its own MDQ direction-setting, I don't think any of us
>>>> would motivate it.
>>> Well, I just found out from Tom last week that TSG is
>>> planning an upgrade to the Ann Arbor network, in part
>>> motivated by the "storm" of network traffic that our
>>> metadata updates cause when clients fetch a new aggregate
>>> each business day. In that regard, trying out a CDN would
>>> probably help alleviate our demand on at least that one
>>> network. It would also gain us operational experience
>>> working with a CDN service. If I had my 'drothers, I would
>>> have two new ops staff already hired who would be part of
>>> this and gain operational knowledge from it. I don't think
>>> that will happen until later this year or next, so if IJ and
>>> others in TSG think they might be able to get some useful
>>> experience out of this, and that it would help with the
>>> network issue or in other ways, I'd be in support.
>> IJ?
>>
>> Can you ask others in TSG?
>>
> I had a chance to talk about it with Mike who is the director for TSG
> and he was very supportive running it on a CDN if that's the
> recommendation.
>
> --IJ
I would also be very supportive of gaining experience with distribution
via a CDN.
--
Nick
Attachment:
signature.asc
Description: OpenPGP digital signature
- Re: [Per-Entity] distribution of aggregate metadata, (continued)
- Re: [Per-Entity] distribution of aggregate metadata, Thomas Lenggenhager, 08/12/2016
- Re: [Per-Entity] distribution of aggregate metadata, Thomas Lenggenhager, 08/12/2016
- Re: [Per-Entity] distribution of aggregate metadata, IJ Kim, 08/12/2016
- Re: [Per-Entity] distribution of aggregate metadata, Cantor, Scott, 08/12/2016
- Re: [Per-Entity] distribution of aggregate metadata, Tom Scavo, 08/12/2016
- Re: [Per-Entity] distribution of aggregate metadata, Nicholas Roy, 08/12/2016
- Re: [Per-Entity] distribution of aggregate metadata, Nicholas Roy, 08/12/2016
- Re: [Per-Entity] distribution of aggregate metadata, Nick Roy, 08/11/2016
- Re: [Per-Entity] distribution of aggregate metadata, IJ Kim, 08/12/2016
- Re: [Per-Entity] distribution of aggregate metadata, Nicholas Roy, 08/12/2016
- Re: [Per-Entity] distribution of aggregate metadata, Tom Scavo, 08/12/2016
- Re: [Per-Entity] distribution of aggregate metadata, Nicholas Roy, 08/12/2016
Archive powered by MHonArc 2.6.19.