On Fri, 15 Nov 2013 00:23:55 +0000 Wookey <woo...@wookware.org> wrote: > Arm64 kit will be even more expensive/unobtanium for a while. > Although I expect the situation to be very different laster next year.
We can schedule a reevaluation of getting server class hardware next year then. Until then however, we still have to fix the current situation. > It's expensive, but not that bad. Or are you talking about 64-bit > hardware? Has anyoneannounced prices for that yet? No that was the price estimate I heard from Hector a while back for the Calxeda boards (is that the highbank ones that Arnaud mentioned? probably but not sure). I'm pretty sure that was the price, but I might be wrong as to what that price involved (ie chassis with boards or single board). Anyway it's expensive, and it's still the matter of redundancy, you can't just buy one. > I think that's nonsense. If we had a box we could find somewhere to > house it. Fine then, I still think the situation would be more complicated, but I'll take your word for it that it's a non-issue. > There was very strong resistance in the room to the use of buildds > without debian kernel support, because it's a major hassle and > security risk. That rules out odroid for the time being. That's fine, I'm not the one to do the security upgrades so I can't possibly insist on putting more burden on other people's shoulders. I'd fancy the idea of super fast builders, but not at the cost of other people doing the grunt work for one port's convenience. > Or Wanda or the Nitrogen6x we've just kindly been offered. Or those, yes. One of the issues is getting boards with more RAM than the current 1GB that the MX53 Locos have, so only the wandboard quad and the utilite fall in that criteria, I see that the Nitrogen6X has only 1GB RAM, so even if it's a nice and tempting offer, I'd still think we need to evaluate this better. It would definitely improve the situation immensely, but builds might still break due to OutOfMemory errors (like iceweasel), but we would definitely get rid of tons of timeout errors that currently need special casing. > What's the mainline status of those? I was going to answer I don't really know, but I just saw that Arnaud just answered that, thanks :) > I think we need to bash out some criteria for deciding during the > mini-debconf. i.e deciding how we decide (or just make a decision if > possible). Better to decide now rather than wait for the perfect solution next year (which might have other problems other than price). Better is Good's worst enemy. > e.g hold out for server-grade kit for a while - if so how long? (0 > months, 3 months, 6months, longer?) If it was a vote, I'd say get some cheap hardware now to replace the current builders, say costing under an absolute limit of 1000GBP (total: disks, boards, cases, PSU, multiserial terminals, etc). That might even get lower if we accept the Nitrogen6X offer. And reevaluate the situation in 12 months when arm server class becomes available (arm64 or arm32). > Is debian kernel an absolute requirement, or are we prepared to risk a > custom kernel if we think it'll only be for 6 months? If DSA absolutely requires kernel support then I don't think there is much we can do. And I don't think that's a promise anyone could actually make, that we expect mainline support to be fixed in the next 6 months. It might take a year or more for that matter and I wouldn't want to be the one to tell DSA "just a couple of weeks more!" :) Though to be honest, I think the main issues, disk, network and SATA should already be working, iMX6 is pretty well supported currently, afaik. > Rate the other options. (Sponsored kit immediately has a significant > advantage :-) > > Markos- as driver on this, could you summarise the relevant features > of odroid, N6x, wanda, utilite?) > > Speed, specs, cost, remote serial, remote power cycle, mounting, > kernel status, availability, power consumption. Anything else > important. I'll prepare a wiki page and post here soon. Regards Konstantinos
pgpNmoEqtgwZ2.pgp
Description: PGP signature
