Re: [sagemath-admins] Re: [sage-devel] Re: *.math.washington.edu hardware resources
Le mercredi 16 septembre 2015 19:13:42 UTC+2, William a écrit : Let's plan to have a page describing the compute resources and > acknowledging *everybody* (or organization) that contributes to having > bought them, hosting them, and helping to run them. > > William > We could use http://wiki.sagemath.org/Infrastructure for that purpose. -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
Re: [sage-devel] Re: *.math.washington.edu hardware resources
On Thu, Sep 17, 2015 at 6:05 PM, Vincent Delecroix <20100.delecr...@gmail.com> wrote: > > > On 17/09/15 17:10, William Stein wrote: >> >> On Thu, Sep 17, 2015 at 1:04 PM, kcrisman wrote: > it would be useful to revive the discussion of a > true SageMath Foundation, separate from SMCI of course, whether > incorporated > in the US or Europe (or elsewhere), but it sounds like there wasn't > currently enough interest in helping with the substantial paperwork > involved > in this as of yet. But if this existed it might help support some of > the > other Sage infrastructure needs, beyond the *.math.washington.edu > resources > this thread was supposed to be about :) Such a foundation isn't going to happen any time soon, at least not with me the one doing it. I talked extensively with a partner at Fenwick & West, and was advised that it is currently basically impossible for an "Open Source Software Not for profit" to get 501c3 status these days in the US. The IRS used to look favorably on such >>> >>> Wow, that is really depressing. It would be interesting to poll various >>> US-based OSS organizations that have such foundations (if any exist) to >>> see >>> how long ago they got such status. Certainly being on a mailing list for >>> UW >>> is preferable to no option for donating at all! >> >> >> You should be able to get removed from that list -- and I will try to >> make it so they remove everybody that was added due to donating to >> Sage Foundation. > > > Sorry for the naive question: why such fondation needs to be based in US? I'm not a lawyer, but I think the main point is for donors to be able to get tax benefits for donating. For US donors, this means that the foundation needs to be registered with the IRS as a nonprofit of some kind (501c3 or otherwise). For European donors, the requirements vary somewhat by country. There's a tradeoff between the paperwork, time and money needed to register with each country and the amount you expect to receive from residents of that country. Since the US has a larger population than any single European country, that's why the discussion usually focuses on it as a US foundation. Of course, having a French, English or German organization would be great too, but requires work. David > > Vincent > > > -- > You received this message because you are subscribed to the Google Groups > "sage-devel" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to sage-devel+unsubscr...@googlegroups.com. > To post to this group, send email to sage-devel@googlegroups.com. > Visit this group at http://groups.google.com/group/sage-devel. > For more options, visit https://groups.google.com/d/optout. -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
Re: [sage-devel] Re: *.math.washington.edu hardware resources
On 17/09/15 17:10, William Stein wrote: On Thu, Sep 17, 2015 at 1:04 PM, kcrisman wrote: it would be useful to revive the discussion of a true SageMath Foundation, separate from SMCI of course, whether incorporated in the US or Europe (or elsewhere), but it sounds like there wasn't currently enough interest in helping with the substantial paperwork involved in this as of yet. But if this existed it might help support some of the other Sage infrastructure needs, beyond the *.math.washington.edu resources this thread was supposed to be about :) Such a foundation isn't going to happen any time soon, at least not with me the one doing it. I talked extensively with a partner at Fenwick & West, and was advised that it is currently basically impossible for an "Open Source Software Not for profit" to get 501c3 status these days in the US. The IRS used to look favorably on such Wow, that is really depressing. It would be interesting to poll various US-based OSS organizations that have such foundations (if any exist) to see how long ago they got such status. Certainly being on a mailing list for UW is preferable to no option for donating at all! You should be able to get removed from that list -- and I will try to make it so they remove everybody that was added due to donating to Sage Foundation. Sorry for the naive question: why such fondation needs to be based in US? Vincent -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
Re: [sage-devel] Re: *.math.washington.edu hardware resources
On Thu, Sep 17, 2015 at 1:04 PM, kcrisman wrote: >> >> > it would be useful to revive the discussion of a >> > true SageMath Foundation, separate from SMCI of course, whether >> > incorporated >> > in the US or Europe (or elsewhere), but it sounds like there wasn't >> > currently enough interest in helping with the substantial paperwork >> > involved >> > in this as of yet. But if this existed it might help support some of >> > the >> > other Sage infrastructure needs, beyond the *.math.washington.edu >> > resources >> > this thread was supposed to be about :) >> >> Such a foundation isn't going to happen any time soon, at least not >> with me the one doing it. I talked extensively with a partner at >> Fenwick & West, and was advised that it is currently basically >> impossible for an "Open Source Software Not for profit" to get 501c3 >> status these days in the US. The IRS used to look favorably on such >> > > Wow, that is really depressing. It would be interesting to poll various > US-based OSS organizations that have such foundations (if any exist) to see > how long ago they got such status. Certainly being on a mailing list for UW > is preferable to no option for donating at all! You should be able to get removed from that list -- and I will try to make it so they remove everybody that was added due to donating to Sage Foundation. > > -- > You received this message because you are subscribed to the Google Groups > "sage-devel" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to sage-devel+unsubscr...@googlegroups.com. > To post to this group, send email to sage-devel@googlegroups.com. > Visit this group at http://groups.google.com/group/sage-devel. > For more options, visit https://groups.google.com/d/optout. -- William (http://wstein.org) -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
Re: [sage-devel] Re: *.math.washington.edu hardware resources
> > > > it would be useful to revive the discussion of a > > true SageMath Foundation, separate from SMCI of course, whether > incorporated > > in the US or Europe (or elsewhere), but it sounds like there wasn't > > currently enough interest in helping with the substantial paperwork > involved > > in this as of yet. But if this existed it might help support some of > the > > other Sage infrastructure needs, beyond the *.math.washington.edu > resources > > this thread was supposed to be about :) > > Such a foundation isn't going to happen any time soon, at least not > with me the one doing it. I talked extensively with a partner at > Fenwick & West, and was advised that it is currently basically > impossible for an "Open Source Software Not for profit" to get 501c3 > status these days in the US. The IRS used to look favorably on such > > Wow, that is really depressing. It would be interesting to poll various US-based OSS organizations that have such foundations (if any exist) to see how long ago they got such status. Certainly being on a mailing list for UW is preferable to no option for donating at all! -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
[sage-devel] Re: *.math.washington.edu hardware resources
I'm not an expert with configuring a cluster, but I've read good things about "Ansible". Do we have any ansible "experts" among us? I think, this tool is a good fit because it describes "goals" and it's extensible with custom modules (written in Python). My uneducated feeling is that we should really look into it … -- harald On Mon, Sep 14, 2015 at 10:13 PM, William Stein wrote: > On Mon, Sep 14, 2015 at 1:10 PM, Volker Braun > wrote: > > I went to a talk by the TravisCI CEO a while ago and one advice that > struck > > a chord was: If you have to log into your server you are doing it wrong. > Its > > of course OK for a one-off machine to debug / play around with, but it > just > > makes administration a time sink. Its better to just erase an instance > and > > recreate it from your orchestration tool. Solves the problem of > documenting > > *what* you changed as a side effect. > > > > On that note, why not openstack / openshift / kubernetes / ...? Of course > > you can create and manage kvm images and/or containers by hand, but why > > reinvent the wheel. > > Let's start fresh! I'm open to any an all ideas for how to setup the > cluster, and *greatly* appreciated your suggestions. > > -- William > > > > > > > -- > > -- > > You received this message because you are subscribed to the Google > > Groups "sage.math users" group. > > To post to this group, send email to sagemath-us...@googlegroups.com > > To unsubscribe from this group, send email to > > sagemath-users+unsubscr...@googlegroups.com > > For more options, visit this group at > > http://groups.google.com/group/sagemath-users > > > > --- > > You received this message because you are subscribed to the Google Groups > > "sage.math users" group. > > To unsubscribe from this group and stop receiving emails from it, send an > > email to sagemath-users+unsubscr...@googlegroups.com. > > > > For more options, visit https://groups.google.com/d/optout. > > > > -- > William (http://wstein.org) > -- Harald Schilly -- SageMath, Inc. https://cloud.sagemath.com Please sign up for a $7/month SageMathCloud membership. Goto Billing tab in Account -> Credit Card -> Add a subscription -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
Re: [sage-devel] Re: *.math.washington.edu hardware resources
On Thu, Sep 17, 2015 at 7:34 AM, kcrisman wrote: >> >> What the hell. You too? It's even worse than I thought, basically >> everybody here has had a share of ODK's money and is thankful for it. >> > > Ahem - definitely not getting ODK money here :) nor are probably 95% (?) of > Sage developers. But given that there seem to be few or no strings attached > to it other than making Sage play more nicely with other mathematical OSS, I > guess people are allowed to be thankful? I'm thankful that people who really > know how to code well will be able to do those things. If I understand > correctly, ODK is way bigger than Sage, also, so I'm thankful that GAP and > other programs that provide much of the most technical mathematical > functionality in many areas of Sage will be able to get better and work > better together with Sage. It is unfortunate that there aren't other > funding sources as well, but as we have seen that is just the state of > things. > > Along those lines, whenever William is able to come up for air from SMC > (perhaps the new year?) Unfortunately, it'll only get worse then, since I'll have to do both running SMC, developing SMC, *and* teaching. At least the Joint Math meetings are a mile from my house this time :-) > it would be useful to revive the discussion of a > true SageMath Foundation, separate from SMCI of course, whether incorporated > in the US or Europe (or elsewhere), but it sounds like there wasn't > currently enough interest in helping with the substantial paperwork involved > in this as of yet. But if this existed it might help support some of the > other Sage infrastructure needs, beyond the *.math.washington.edu resources > this thread was supposed to be about :) Such a foundation isn't going to happen any time soon, at least not with me the one doing it. I talked extensively with a partner at Fenwick & West, and was advised that it is currently basically impossible for an "Open Source Software Not for profit" to get 501c3 status these days in the US. The IRS used to look favorably on such things a few years ago (e.g., Mozilla!), but now looks very unfavorably on open source as being "not for profit" (things that are political can change dramatically from one point in time to another). I didn't believe this could be the case, but the Fenwick & West partner is a top expert who is very familiar with what is going on, and she repeatedly clarified that it is. She suggested starting a 501c3 that has nothing to do with Sage, and also doesn't have the word "Sage" in its title would be the only reasonable strategy... and maybe later it would get involved with supporting Sage. Even doing that, the work involved is more than starting a company, and the rules are tricky involving taxes, so I definitely don't have the time to do that now. For all it's cons, University of Washington does at least administer the "Sage Foundation" budget very professionally, they don't charge anything in overhead (absolutely 100% of donations are spent on sage activities with no cut at all), and they take care of all the dispersement of funds to people (e.g., participants of "Women in Sage" Sage days -- often funded from this, bits of hardware we need for a conference, etc.).One drawback is their stupid "please donate again" mailing list. - William -- William (http://wstein.org) -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
Re: [sage-devel] Re: *.math.washington.edu hardware resources
> > > What the hell. You too? It's even worse than I thought, basically > everybody here has had a share of ODK's money and is thankful for it. > > Ahem - definitely not getting ODK money here :) nor are probably 95% (?) of Sage developers. But given that there seem to be few or no strings attached to it other than making Sage play more nicely with other mathematical OSS, I guess people are allowed to be thankful? I'm thankful that people who really know how to code well will be able to do those things. If I understand correctly, ODK is way bigger than Sage, also, so I'm thankful that GAP and other programs that provide much of the most technical mathematical functionality in many areas of Sage will be able to get better and work better together with Sage. It is unfortunate that there aren't other funding sources as well, but as we have seen that is just the state of things. Along those lines, whenever William is able to come up for air from SMC (perhaps the new year?) it would be useful to revive the discussion of a true SageMath Foundation, separate from SMCI of course, whether incorporated in the US or Europe (or elsewhere), but it sounds like there wasn't currently enough interest in helping with the substantial paperwork involved in this as of yet. But if this existed it might help support some of the other Sage infrastructure needs, beyond the *.math.washington.edu resources this thread was supposed to be about :) -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
Re: [sage-devel] Re: *.math.washington.edu hardware resources
On 17 September 2015 at 10:26, Nathann Cohen wrote: >> that grant (e.g. for me, 3 person-months over 4 years which translates >> into about 2 days per week) are required to keep time sheets so that That should read: 2 days per month (it is 1/16). > > What the hell. You too? It's even worse than I thought, basically > everybody here has had a share of ODK's money and is thankful for it. It's not such a big deal, just one signature on a piece of paper (or electronic) per month -- and in my case the signature will be that of my head of department who certainly will not ask me just what I do all day. John > > Nathann -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
Re: [sage-devel] Re: *.math.washington.edu hardware resources
> that grant (e.g. for me, 3 person-months over 4 years which translates > into about 2 days per week) are required to keep time sheets so that What the hell. You too? It's even worse than I thought, basically everybody here has had a share of ODK's money and is thankful for it. Nathann -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
Re: [sage-devel] Re: *.math.washington.edu hardware resources
On 16 September 2015 at 19:03, Dima Pasechnik wrote: > On 16 September 2015 at 10:36, Nathann Cohen wrote: >> Dima, >> >>> E.g. as far as I am concerned, it's thanks to ODK, which pays 50% of my >>> salary, that I can set aside time for this. >> >> You are of course welcome (and probably bound) to spend 50% of your working >> time >> on whatever is in OpenDreamKit's plan. >> >> Now, I'm sorry to say that I read Nicolas' message as an accountant's >> attempt to >> try to fit in his report something that has not even been done yet, and does >> not >> seem (according to him) to clearly enter its scope. Which is why I got the >> impression that he planned to stamp 'ODK' on whatever happened to pass by, >> >> to claim later that it had been done 'thanks to ODK'. > > No, your impression is wrong. It is an unfortunate reality that EU > (and other) grants come with all sorts of bells and whistles attached, > and one needs accounting tricks to bend the rules they set up to spend > money on what's important. And maintaining UW hardware for Sage is > important for success of ODK, there is no question about it. > > If ODK did not materialize, there is no telling what sorts of > implications it would have on Sage as a project, but surely nothing > positive. It is a legal requirement for work supported financially by the ODK grant -- which definitely include's Dima's 50% time -- to acknowledge that support. Moreover Dima and others whose time is supported even in small part by that grant (e.g. for me, 3 person-months over 4 years which translates into about 2 days per week) are required to keep time sheets so that an auditor can be satisfied that the money paid by the EC for that timew has been spent on what it was provided for. Yes, it is very bureaucratic but the benefits outweight that! John > > Please try to understand, once again, that you are at CNRS in a very > fortunate and exceptional position, and many people are much less > lucky. Try to get a grant and manage it to see my point :-) > > Dima > >> >> I'm sure that there are enough original ideas in ODK's project to keep >> everybody >> busy for 5 years, and I surely won't blame you for working hard on them. >> >> Nathann >> >> P.S.: nothing indeed prevents you from working on topics unrelated to ODK >> during >> the other 50% of your time, e.g. I am not paid to write Sage code. > > -- > -- > You received this message because you are subscribed to the Google > Groups "sage.math users" group. > To post to this group, send email to sagemath-us...@googlegroups.com > To unsubscribe from this group, send email to > sagemath-users+unsubscr...@googlegroups.com > For more options, visit this group at > http://groups.google.com/group/sagemath-users > > --- > You received this message because you are subscribed to the Google Groups > "sage.math users" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to sagemath-users+unsubscr...@googlegroups.com. > For more options, visit https://groups.google.com/d/optout. -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
Re: [sage-devel] Re: *.math.washington.edu hardware resources
I just updated the project with some excel + csv files with a much more indepth inventory. Also, concerning ansible -- this is what I used when working on things this summer (see https://github.com/ohanar/sagemath.org). I found that when you use any well supported plugin, using ansible was very pleasant, but once you stepped outside of the ecosystem -- it got to be a pain very quickly. A lot of this can be solved by writing your own plugins, but I found that to also have quite a learning curve (ansible does some python magic that defines various global variables in your modules). An alternative to look at is saltstack -- it is also written in python, and also uses yaml as its primary configuration format. I've heard writing plugins for it is simpler as well, but I haven't investigated it thoroughly. On Wed, Sep 16, 2015 at 2:08 PM, kcrisman wrote: > > > Please try to understand, once again, that you are at CNRS in a very >> fortunate and exceptional position, and many people are much less >> lucky. Try to get a grant and manage it to see my point :-) >> >> >> > And (not speaking to Nathann or anyone else) try it when teaching a 3-3 or > 4-4 load at a place where teaching by far is the main component to > tenure... I am hoping that more and more of them will do like Greg Bard did > and write a Sage teaching book, but that will take time (and maybe MBX!). > But for developing it's a real challenge to find that reliable time. > > -- > You received this message because you are subscribed to the Google Groups > "sage-devel" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to sage-devel+unsubscr...@googlegroups.com. > To post to this group, send email to sage-devel@googlegroups.com. > Visit this group at http://groups.google.com/group/sage-devel. > For more options, visit https://groups.google.com/d/optout. > -- Andrew -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
Re: [sage-devel] Re: *.math.washington.edu hardware resources
Please try to understand, once again, that you are at CNRS in a very > fortunate and exceptional position, and many people are much less > lucky. Try to get a grant and manage it to see my point :-) > > > And (not speaking to Nathann or anyone else) try it when teaching a 3-3 or 4-4 load at a place where teaching by far is the main component to tenure... I am hoping that more and more of them will do like Greg Bard did and write a Sage teaching book, but that will take time (and maybe MBX!). But for developing it's a real challenge to find that reliable time. -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
Re: [sage-devel] Re: *.math.washington.edu hardware resources
On Wed, Sep 16, 2015 at 07:23:26AM -0700, William Stein wrote: > Related question: Does the ODK grant have a hardware or computing > budget or money that could be allocated to that? If so, how are you > planning to use it (or is there a plan yet)? I ask only because > we might coordinate our plans to minimize effort and possibly improve > backups/redundancy. Buying hardware on this type of grant is not very practical: IIRC, it would only pay for the four first years of usage of the hardware and thus requires for some cofunding. So we did not plan much in the direct costs: looking back at the proposal, it's planned to buy servers in Bremen, Grenoble, Saint Andrews, SouthHampton, Sheffield. We might be able to use some of the indirect costs for buying more hardware besides laptops and the like. However I am not yet familiar with how everything works, so I'd rather wait a bit before saying more. In any cases, it would need to be hosted "locally". Does it sound technically possible to use non-local hardware for the "sage farm" with little additional sysadmin overheard? If yes, what would it take, in term of bandwidth as well as technical and legal permissions on the host network? Cheers, Nicolas -- Nicolas M. Thiéry "Isil" http://Nicolas.Thiery.name/ -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
Re: [sage-devel] Re: *.math.washington.edu hardware resources
On 16 September 2015 at 10:36, Nathann Cohen wrote: > Dima, > >> E.g. as far as I am concerned, it's thanks to ODK, which pays 50% of my >> salary, that I can set aside time for this. > > You are of course welcome (and probably bound) to spend 50% of your working > time > on whatever is in OpenDreamKit's plan. > > Now, I'm sorry to say that I read Nicolas' message as an accountant's > attempt to > try to fit in his report something that has not even been done yet, and does > not > seem (according to him) to clearly enter its scope. Which is why I got the > impression that he planned to stamp 'ODK' on whatever happened to pass by, > > to claim later that it had been done 'thanks to ODK'. No, your impression is wrong. It is an unfortunate reality that EU (and other) grants come with all sorts of bells and whistles attached, and one needs accounting tricks to bend the rules they set up to spend money on what's important. And maintaining UW hardware for Sage is important for success of ODK, there is no question about it. If ODK did not materialize, there is no telling what sorts of implications it would have on Sage as a project, but surely nothing positive. Please try to understand, once again, that you are at CNRS in a very fortunate and exceptional position, and many people are much less lucky. Try to get a grant and manage it to see my point :-) Dima > > I'm sure that there are enough original ideas in ODK's project to keep > everybody > busy for 5 years, and I surely won't blame you for working hard on them. > > Nathann > > P.S.: nothing indeed prevents you from working on topics unrelated to ODK > during > the other 50% of your time, e.g. I am not paid to write Sage code. -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
Re: [sage-devel] Re: *.math.washington.edu hardware resources
Dima, > E.g. as far as I am concerned, it's thanks to ODK, which pays 50% of my > salary, that I can set aside time for this. You are of course welcome (and probably bound) to spend 50% of your working time on whatever is in OpenDreamKit's plan. Now, I'm sorry to say that I read Nicolas' message as an accountant's attempt to try to fit in his report something that has not even been done yet, and does not seem (according to him) to clearly enter its scope. Which is why I got the impression that he planned to stamp 'ODK' on whatever happened to pass by, and to claim later that it had been done 'thanks to ODK'. I'm sure that there are enough original ideas in ODK's project to keep everybody busy for 5 years, and I surely won't blame you for working hard on them. Nathann P.S.: nothing indeed prevents you from working on topics unrelated to ODK during the other 50% of your time, e.g. I am not paid to write Sage code. -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
Re: [sagemath-admins] Re: [sage-devel] Re: *.math.washington.edu hardware resources
On Wed, Sep 16, 2015 at 9:48 AM, Jan Groenewald wrote: > Hi > > On 16 September 2015 at 18:44, Dima Pasechnik wrote: >> >> >> Nicolas is totally correct here. E.g. as far as I am concerned, it's >> thanks to ODK, which pays 50% of my salary, that I can set aside time >> for this. >> > > > I work on this because of when it aligns with AIMS goals/needs (some of > which I set) of using Sagemath for teaching and research. Similar feeling > from me; I don't particularly need an AIMS stamp on anything, though it is > nice to recognize the contribution. > Let's plan to have a page describing the compute resources and acknowledging *everybody* (or organization) that contributes to having bought them, hosting them, and helping to run them. William > Regards, > Jan > > > > -- > .~. > /V\ Jan Groenewald > /( )\www.aims.ac.za > ^^-^^ > > -- > > --- > You received this message because you are subscribed to the Google Groups > "sagemath-admins" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to sagemath-admins+unsubscr...@googlegroups.com. > > For more options, visit https://groups.google.com/d/optout. -- William (http://wstein.org) -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
Re: [sagemath-admins] Re: [sage-devel] Re: *.math.washington.edu hardware resources
Hi On 16 September 2015 at 18:44, Dima Pasechnik wrote: > > Nicolas is totally correct here. E.g. as far as I am concerned, it's > thanks to ODK, which pays 50% of my salary, that I can set aside time > for this. > > I work on this because of when it aligns with AIMS goals/needs (some of which I set) of using Sagemath for teaching and research. Similar feeling from me; I don't particularly need an AIMS stamp on anything, though it is nice to recognize the contribution. Regards, Jan -- .~. /V\ Jan Groenewald /( )\www.aims.ac.za ^^-^^ -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
Re: [sage-devel] Re: *.math.washington.edu hardware resources
On 16 September 2015 at 02:40, Nathann Cohen wrote: >> I believe this will help quite some ODK toward its aims. So, Vincent, >> Samuel, Dima, if you feel like it, it would be natural to count some >> of your work here as part of your ODK contribution. We need to fit this >> into some of our deliverables. The closest match I found for now is: >> >>D3.8: Continuous integration platform for multi-platform build/test. >> >> Not great; better ideas? > > > Cool. So basically you add a "ODK" stamp over what people would do anyway, > and then you say that it has been done thanks to ODK? Nathann, Nicolas is totally correct here. E.g. as far as I am concerned, it's thanks to ODK, which pays 50% of my salary, that I can set aside time for this. If I had been on a permanent (non)tenure-track in some reincarnation of State U. of Mordor, or had to write code for food, or had to teach 15 hours a week, I most probably won't be here. Cheers, Dima > > As far I am concerned, I don't mind if you manage your small company on > another mailing list. > > Nathann -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
Re: [sage-devel] Re: *.math.washington.edu hardware resources
On Wed, Sep 16, 2015 at 7:25 AM, William Stein wrote: > Hi, > > A concrete first task is to add everybody on the following list to the > sagemath-admins mailing list (by tracking down their emails), etc.. Samuel Lelievre has volunteered to do this. Thanks! > Can one of you volunteer to do this, and once you do I'll add you and > make you an owner of that list, so you can add other people: > > - William Stein > - Andrey Novoseltsev > - Dima Pasechnik > - Vincent Delecroix > - Thierry > - Samuel Lelievre > - Harald Schilly > - Jan Groenewald > - Andrew Ohana (minimal availability -- but wants to discuss) > - Keith Clawson (very minimal availability, but is local and has > **extremely** good hardware skills) > > On Wed, Sep 16, 2015 at 7:23 AM, William Stein wrote: >> On Wed, Sep 16, 2015 at 2:36 AM, Nicolas M. Thiery >> wrote: >>> On Tue, Sep 15, 2015 at 11:55:55AM -0700, William Stein wrote: I think the current Sage UW cluster team is: - William Stein - Andrey Novoseltsev - Dima Pasechnik - Vincent Delecroix - Thierry - Samuel Lelievre - Harald Schilly - Andrew Ohana (minimal availability -- but wants to discuss) >>> >>> Thanks everyone! >>> >>> I believe this will help quite some ODK toward its aims. So, Vincent, >>> Samuel, Dima, if you feel like it, it would be natural to count some >>> of your work here as part of your ODK contribution. We need to fit this >>> into some of our deliverables. The closest match I found for now is: >>> >>>D3.8: Continuous integration platform for multi-platform build/test. >>> >>> Not great; better ideas? >> >> Related question: Does the ODK grant have a hardware or computing >> budget or money that could be allocated to that? If so, how are you >> planning to use it (or is there a plan yet)? I ask only because >> we might coordinate our plans to minimize effort and possibly improve >> backups/redundancy. >> >> William >> >>> >>> Cheers, >>> Nicolas >>> -- >>> Nicolas M. Thiéry "Isil" >>> http://Nicolas.Thiery.name/ >>> >>> -- >>> -- >>> You received this message because you are subscribed to the Google >>> Groups "sage.math users" group. >>> To post to this group, send email to sagemath-us...@googlegroups.com >>> To unsubscribe from this group, send email to >>> sagemath-users+unsubscr...@googlegroups.com >>> For more options, visit this group at >>> http://groups.google.com/group/sagemath-users >>> >>> --- >>> You received this message because you are subscribed to the Google Groups >>> "sage.math users" group. >>> To unsubscribe from this group and stop receiving emails from it, send an >>> email to sagemath-users+unsubscr...@googlegroups.com. >>> For more options, visit https://groups.google.com/d/optout. >> >> >> >> -- >> William (http://wstein.org) > > > > -- > William (http://wstein.org) -- William (http://wstein.org) -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
Re: [sage-devel] Re: *.math.washington.edu hardware resources
On Wed, Sep 16, 2015 at 8:09 AM, kcrisman wrote: > > >>> As far I am concerned, I don't mind if you manage your small company on >>> another mailing list. >>> >>> >> Surely you cared way more about Sage succeeding than petty issues of >> credit!? >> > > > Just to be clear, it does seem a bit confusing at times what discussions are > SMC Inc related and [...] Nathann wrote to me to clarify what "company" he was talking about above. It has nothing to do with SageMath, Inc. or SMC. He was referring to ODK as "a small company". William -- William (http://wstein.org) -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
[sage-devel] Re: *.math.washington.edu hardware resources
As far I am concerned, I don't mind if you manage your small company on >> another mailing list. >> >> >> Surely you cared way more about Sage succeeding than petty issues of > credit!? > > Just to be clear, it does seem a bit confusing at times what discussions are SMC Inc related and what discussions are Sage-related, and presumably that will remain so for a while. In this case I think it was pretty clear that the servers in question were not SMC related at all. But I agree that it would be nice to have some separation there. As a practical example of this, suppose that one wanted to donate funds to Sage, but did *not* want to subsidize SMC (or maybe wanted a tax deductible donation, or whatever). What activities would Sage Foundation donations (in its current instantiation as a UW fund) support, which ones would other things support? E.g. it sounds like the Sage booth this year at the Joint Meetings is by SMC, not Sage per se, though in the past the UW fund supported the Sage booth (as far as I recall). So maybe the distinction should be teased out a bit in emails - not that SMC things couldn't be discussed here, after all the software is GPL, but still I hope you see my point that the boundary occasionally can be unclear. - kcrisman -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
Re: [sage-devel] Re: *.math.washington.edu hardware resources
Hi On 16 September 2015 at 16:20, William Stein wrote: > Hi, > > There's starting to be some initial work in > > > https://cloud.sagemath.com/projects/f125ea34-9817-4715-b879-e26e0c7fcb7c/files/public/ > > Who on the list of people below has experience with managing a > nontrivial hardware cluster and wants to share how they do it? Jan > Groenewald, maybe you can say more? > We're getting quite comfortable on Ubuntu or Debian running LXC containers. We're in progress of replacing backup scripts with backuppc, and we're interested to move to ansible (or another) as soon as we have time. If everyone else here is more of a mathematician / programmer than sysadm, then yes, 13 years ago I moved from math to sysadm, and that is our focus. I must confess our availability is not good at all, and a main goal for us would be to move buildslaves and PPA autobuild LXC containers (and their backups) from our Hetzner.de to some of these machines, to free up CPU for our institution. Our total hardware is less than you are talking about here. We'd be more than happy to continue to maintain and to document those, and over time might become more available for other tasks, and we're certainly interested in following. It would be a doable task to make available containers for anyone who wanted to maintain a container buildslave for any other Linux distro. Perhaps one host could be dedicated for that? We'd also be very interested in learning from any implemented cluster architecture, configuration management, and have some very minimal use of ansible. Regards, Jan -- .~. /V\ Jan Groenewald /( )\www.aims.ac.za ^^-^^ -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
Re: [sage-devel] Re: *.math.washington.edu hardware resources
On Wed, Sep 16, 2015 at 7:35 AM, Jean-Pierre Flori wrote: > Hi all, > > I can also help if needed. Thanks! I've added you to the project and list: https://cloud.sagemath.com/projects/f125ea34-9817-4715-b879-e26e0c7fcb7c/files/public/ > Maybe that would make the french admins count too high :) > > Best, > JP > > On Wednesday, September 16, 2015 at 4:26:08 PM UTC+2, William wrote: >> >> Hi, >> >> A concrete first task is to add everybody on the following list to the >> sagemath-admins mailing list (by tracking down their emails), etc.. >> Can one of you volunteer to do this, and once you do I'll add you and >> make you an owner of that list, so you can add other people: >> >> - William Stein >> - Andrey Novoseltsev >> - Dima Pasechnik >> - Vincent Delecroix >> - Thierry >> - Samuel Lelievre >> - Harald Schilly >> - Jan Groenewald >> - Andrew Ohana (minimal availability -- but wants to discuss) >> - Keith Clawson (very minimal availability, but is local and has >> **extremely** good hardware skills) >> >> On Wed, Sep 16, 2015 at 7:23 AM, William Stein wrote: >> > On Wed, Sep 16, 2015 at 2:36 AM, Nicolas M. Thiery >> > wrote: >> >> On Tue, Sep 15, 2015 at 11:55:55AM -0700, William Stein wrote: >> >>> I think the current Sage UW cluster team is: >> >>> >> >>>- William Stein >> >>>- Andrey Novoseltsev >> >>>- Dima Pasechnik >> >>>- Vincent Delecroix >> >>>- Thierry >> >>>- Samuel Lelievre >> >>>- Harald Schilly >> >>>- Andrew Ohana (minimal availability -- but wants to discuss) >> >> >> >> Thanks everyone! >> >> >> >> I believe this will help quite some ODK toward its aims. So, Vincent, >> >> Samuel, Dima, if you feel like it, it would be natural to count some >> >> of your work here as part of your ODK contribution. We need to fit this >> >> into some of our deliverables. The closest match I found for now is: >> >> >> >>D3.8: Continuous integration platform for multi-platform build/test. >> >> >> >> Not great; better ideas? >> > >> > Related question: Does the ODK grant have a hardware or computing >> > budget or money that could be allocated to that? If so, how are you >> > planning to use it (or is there a plan yet)? I ask only because >> > we might coordinate our plans to minimize effort and possibly improve >> > backups/redundancy. >> > >> > William >> > >> >> >> >> Cheers, >> >> Nicolas >> >> -- >> >> Nicolas M. Thiéry "Isil" >> >> http://Nicolas.Thiery.name/ >> >> >> >> -- >> >> -- >> >> You received this message because you are subscribed to the Google >> >> Groups "sage.math users" group. >> >> To post to this group, send email to sagemat...@googlegroups.com >> >> To unsubscribe from this group, send email to >> >> sagemath-user...@googlegroups.com >> >> For more options, visit this group at >> >> http://groups.google.com/group/sagemath-users >> >> >> >> --- >> >> You received this message because you are subscribed to the Google >> >> Groups "sage.math users" group. >> >> To unsubscribe from this group and stop receiving emails from it, send >> >> an email to sagemath-user...@googlegroups.com. >> >> For more options, visit https://groups.google.com/d/optout. >> > >> > >> > >> > -- >> > William (http://wstein.org) >> >> >> >> -- >> William (http://wstein.org) > > -- > You received this message because you are subscribed to the Google Groups > "sage-devel" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to sage-devel+unsubscr...@googlegroups.com. > To post to this group, send email to sage-devel@googlegroups.com. > Visit this group at http://groups.google.com/group/sage-devel. > For more options, visit https://groups.google.com/d/optout. -- William (http://wstein.org) -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
Re: [sage-devel] Re: *.math.washington.edu hardware resources
Hi all, I can also help if needed. Maybe that would make the french admins count too high :) Best, JP On Wednesday, September 16, 2015 at 4:26:08 PM UTC+2, William wrote: > > Hi, > > A concrete first task is to add everybody on the following list to the > sagemath-admins mailing list (by tracking down their emails), etc.. > Can one of you volunteer to do this, and once you do I'll add you and > make you an owner of that list, so you can add other people: > > - William Stein > - Andrey Novoseltsev > - Dima Pasechnik > - Vincent Delecroix > - Thierry > - Samuel Lelievre > - Harald Schilly > - Jan Groenewald > - Andrew Ohana (minimal availability -- but wants to discuss) > - Keith Clawson (very minimal availability, but is local and has > **extremely** good hardware skills) > > On Wed, Sep 16, 2015 at 7:23 AM, William Stein > wrote: > > On Wed, Sep 16, 2015 at 2:36 AM, Nicolas M. Thiery > > > wrote: > >> On Tue, Sep 15, 2015 at 11:55:55AM -0700, William Stein wrote: > >>> I think the current Sage UW cluster team is: > >>> > >>>- William Stein > >>>- Andrey Novoseltsev > >>>- Dima Pasechnik > >>>- Vincent Delecroix > >>>- Thierry > >>>- Samuel Lelievre > >>>- Harald Schilly > >>>- Andrew Ohana (minimal availability -- but wants to discuss) > >> > >> Thanks everyone! > >> > >> I believe this will help quite some ODK toward its aims. So, Vincent, > >> Samuel, Dima, if you feel like it, it would be natural to count some > >> of your work here as part of your ODK contribution. We need to fit this > >> into some of our deliverables. The closest match I found for now is: > >> > >>D3.8: Continuous integration platform for multi-platform build/test. > >> > >> Not great; better ideas? > > > > Related question: Does the ODK grant have a hardware or computing > > budget or money that could be allocated to that? If so, how are you > > planning to use it (or is there a plan yet)? I ask only because > > we might coordinate our plans to minimize effort and possibly improve > > backups/redundancy. > > > > William > > > >> > >> Cheers, > >> Nicolas > >> -- > >> Nicolas M. Thiéry "Isil" > > >> http://Nicolas.Thiery.name/ > >> > >> -- > >> -- > >> You received this message because you are subscribed to the Google > >> Groups "sage.math users" group. > >> To post to this group, send email to sagemat...@googlegroups.com > > >> To unsubscribe from this group, send email to > >> sagemath-user...@googlegroups.com > >> For more options, visit this group at > >> http://groups.google.com/group/sagemath-users > >> > >> --- > >> You received this message because you are subscribed to the Google > Groups "sage.math users" group. > >> To unsubscribe from this group and stop receiving emails from it, send > an email to sagemath-user...@googlegroups.com . > >> For more options, visit https://groups.google.com/d/optout. > > > > > > > > -- > > William (http://wstein.org) > > > > -- > William (http://wstein.org) > -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
Re: [sage-devel] Re: *.math.washington.edu hardware resources
Hi, A concrete first task is to add everybody on the following list to the sagemath-admins mailing list (by tracking down their emails), etc.. Can one of you volunteer to do this, and once you do I'll add you and make you an owner of that list, so you can add other people: - William Stein - Andrey Novoseltsev - Dima Pasechnik - Vincent Delecroix - Thierry - Samuel Lelievre - Harald Schilly - Jan Groenewald - Andrew Ohana (minimal availability -- but wants to discuss) - Keith Clawson (very minimal availability, but is local and has **extremely** good hardware skills) On Wed, Sep 16, 2015 at 7:23 AM, William Stein wrote: > On Wed, Sep 16, 2015 at 2:36 AM, Nicolas M. Thiery > wrote: >> On Tue, Sep 15, 2015 at 11:55:55AM -0700, William Stein wrote: >>> I think the current Sage UW cluster team is: >>> >>>- William Stein >>>- Andrey Novoseltsev >>>- Dima Pasechnik >>>- Vincent Delecroix >>>- Thierry >>>- Samuel Lelievre >>>- Harald Schilly >>>- Andrew Ohana (minimal availability -- but wants to discuss) >> >> Thanks everyone! >> >> I believe this will help quite some ODK toward its aims. So, Vincent, >> Samuel, Dima, if you feel like it, it would be natural to count some >> of your work here as part of your ODK contribution. We need to fit this >> into some of our deliverables. The closest match I found for now is: >> >>D3.8: Continuous integration platform for multi-platform build/test. >> >> Not great; better ideas? > > Related question: Does the ODK grant have a hardware or computing > budget or money that could be allocated to that? If so, how are you > planning to use it (or is there a plan yet)? I ask only because > we might coordinate our plans to minimize effort and possibly improve > backups/redundancy. > > William > >> >> Cheers, >> Nicolas >> -- >> Nicolas M. Thiéry "Isil" >> http://Nicolas.Thiery.name/ >> >> -- >> -- >> You received this message because you are subscribed to the Google >> Groups "sage.math users" group. >> To post to this group, send email to sagemath-us...@googlegroups.com >> To unsubscribe from this group, send email to >> sagemath-users+unsubscr...@googlegroups.com >> For more options, visit this group at >> http://groups.google.com/group/sagemath-users >> >> --- >> You received this message because you are subscribed to the Google Groups >> "sage.math users" group. >> To unsubscribe from this group and stop receiving emails from it, send an >> email to sagemath-users+unsubscr...@googlegroups.com. >> For more options, visit https://groups.google.com/d/optout. > > > > -- > William (http://wstein.org) -- William (http://wstein.org) -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
Re: [sage-devel] Re: *.math.washington.edu hardware resources
On Wed, Sep 16, 2015 at 2:36 AM, Nicolas M. Thiery wrote: > On Tue, Sep 15, 2015 at 11:55:55AM -0700, William Stein wrote: >> I think the current Sage UW cluster team is: >> >>- William Stein >>- Andrey Novoseltsev >>- Dima Pasechnik >>- Vincent Delecroix >>- Thierry >>- Samuel Lelievre >>- Harald Schilly >>- Andrew Ohana (minimal availability -- but wants to discuss) > > Thanks everyone! > > I believe this will help quite some ODK toward its aims. So, Vincent, > Samuel, Dima, if you feel like it, it would be natural to count some > of your work here as part of your ODK contribution. We need to fit this > into some of our deliverables. The closest match I found for now is: > >D3.8: Continuous integration platform for multi-platform build/test. > > Not great; better ideas? Related question: Does the ODK grant have a hardware or computing budget or money that could be allocated to that? If so, how are you planning to use it (or is there a plan yet)? I ask only because we might coordinate our plans to minimize effort and possibly improve backups/redundancy. William > > Cheers, > Nicolas > -- > Nicolas M. Thiéry "Isil" > http://Nicolas.Thiery.name/ > > -- > -- > You received this message because you are subscribed to the Google > Groups "sage.math users" group. > To post to this group, send email to sagemath-us...@googlegroups.com > To unsubscribe from this group, send email to > sagemath-users+unsubscr...@googlegroups.com > For more options, visit this group at > http://groups.google.com/group/sagemath-users > > --- > You received this message because you are subscribed to the Google Groups > "sage.math users" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to sagemath-users+unsubscr...@googlegroups.com. > For more options, visit https://groups.google.com/d/optout. -- William (http://wstein.org) -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
Re: [sage-devel] Re: *.math.washington.edu hardware resources
Hi, There's starting to be some initial work in https://cloud.sagemath.com/projects/f125ea34-9817-4715-b879-e26e0c7fcb7c/files/public/ Who on the list of people below has experience with managing a nontrivial hardware cluster and wants to share how they do it? Jan Groenewald, maybe you can say more? - William Stein - Andrey Novoseltsev - Dima Pasechnik - Vincent Delecroix - Thierry - Samuel Lelievre - Harald Schilly - Jan Groenewald - Andrew Ohana (minimal availability -- but wants to discuss) - Keith Clawson (very minimal availability, but is local and has **extremely** good hardware skills) On Wed, Sep 16, 2015 at 6:14 AM, Nathann Cohen wrote: > William, > > If you want to discuss this issue further, please send me a private email. > > Nathann > > -- > You received this message because you are subscribed to the Google Groups > "sage-devel" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to sage-devel+unsubscr...@googlegroups.com. > To post to this group, send email to sage-devel@googlegroups.com. > Visit this group at http://groups.google.com/group/sage-devel. > For more options, visit https://groups.google.com/d/optout. -- William (http://wstein.org) -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
Re: [sage-devel] Re: *.math.washington.edu hardware resources
William, If you want to discuss this issue further, please send me a private email. Nathann -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
[sage-devel] Re: *.math.washington.edu hardware resources
On Wednesday, September 16, 2015, Nathann Cohen wrote: > I believe this will help quite some ODK toward its aims. So, Vincent, >> Samuel, Dima, if you feel like it, it would be natural to count some >> of your work here as part of your ODK contribution. We need to fit this >> into some of our deliverables. The closest match I found for now is: >> >>D3.8: Continuous integration platform for multi-platform build/test. >> >> Not great; better ideas? >> > > Cool. So basically you add a "ODK" stamp over what people would do anyway, > and then you say that it has been done thanks to ODK? > That's not what he said. He said he would use the grant to fund their work. It's a hugely valuable contribution. > As far I am concerned, I don't mind if you manage your small company on > another mailing list. > > Nathann > Surely you cared way more about Sage succeeding than petty issues of credit!? > -- > You received this message because you are subscribed to the Google Groups > "sage-devel" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to sage-devel+unsubscr...@googlegroups.com > > . > To post to this group, send email to sage-devel@googlegroups.com > . > Visit this group at http://groups.google.com/group/sage-devel. > For more options, visit https://groups.google.com/d/optout. > -- Sent from my massive iPhone 6 plus. -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
Re: [sage-devel] Re: *.math.washington.edu hardware resources
> > I believe this will help quite some ODK toward its aims. So, Vincent, > Samuel, Dima, if you feel like it, it would be natural to count some > of your work here as part of your ODK contribution. We need to fit this > into some of our deliverables. The closest match I found for now is: > >D3.8: Continuous integration platform for multi-platform build/test. > > Not great; better ideas? > Cool. So basically you add a "ODK" stamp over what people would do anyway, and then you say that it has been done thanks to ODK? As far I am concerned, I don't mind if you manage your small company on another mailing list. Nathann -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
Re: [sage-devel] Re: *.math.washington.edu hardware resources
On Tue, Sep 15, 2015 at 11:55:55AM -0700, William Stein wrote: > I think the current Sage UW cluster team is: > >- William Stein >- Andrey Novoseltsev >- Dima Pasechnik >- Vincent Delecroix >- Thierry >- Samuel Lelievre >- Harald Schilly >- Andrew Ohana (minimal availability -- but wants to discuss) Thanks everyone! I believe this will help quite some ODK toward its aims. So, Vincent, Samuel, Dima, if you feel like it, it would be natural to count some of your work here as part of your ODK contribution. We need to fit this into some of our deliverables. The closest match I found for now is: D3.8: Continuous integration platform for multi-platform build/test. Not great; better ideas? Cheers, Nicolas -- Nicolas M. Thiéry "Isil" http://Nicolas.Thiery.name/ -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
Re: [sage-devel] Re: *.math.washington.edu hardware resources
Hi William, Please add me too. We'd be happy to help with what we already do: buildslaves for Debian/Ubuntu, PPA autobuild, perhaps some of this can be migrated onto those servers. We're most comfortable working on Debian/Ubuntu with LXC. Regards, Jan On 15 September 2015 at 21:22, William Stein wrote: > Hi, > > I've created a SageMathCloud project that we can use for organization. > I've added everybody in "the team" to that: > > > https://cloud.sagemath.com/projects/f125ea34-9817-4715-b879-e26e0c7fcb7c/files/ > > I've also made a folder "public" > > > https://cloud.sagemath.com/projects/f125ea34-9817-4715-b879-e26e0c7fcb7c/files/public/ > > which is public to the world. > > Anyway, this will be a good place to record schedules, have chat, > layout the topology of the system, record security stuff and > decisions, etc. > > -- William > > On Tue, Sep 15, 2015 at 11:55 AM, William Stein wrote: > > Hi Everybody, > > > > I think the current Sage UW cluster team is: > > > >- William Stein > >- Andrey Novoseltsev > >- Dima Pasechnik > >- Vincent Delecroix > >- Thierry > >- Samuel Lelievre > >- Harald Schilly > >- Andrew Ohana (minimal availability -- but wants to discuss) > > > > I think that's pretty good actually. To be in the team you have to > > volunteer to help out regularly for **the next year**. In exchange > > you get to help out, get admin on everything, and together we'll > > decide on what to do. > > > > Every option is on the table to start. > > > > There were some proposals above already about how to set things up. > > I propose that we come up with ideas and get ideas from the community, > > but that we together have final say on what to actually do, since > > we'll be doing all the work. We can vote/argue, but if there is major > > contention or a tie, I'll decide things somehow. > > > > Our goal (and the legally agreed upon thing with my university) is to > > support (a) sage developers, and (b) research mathematicians. > > > > Regarding hardware, there was no budget when I first sent this email, > > but now there is a small budget due to a much-appreciated (!) donation > > from a Sage developer. So, e.g., buying another full rack is an > > option. > > > > -- William > > > > > > On Tue, Sep 15, 2015 at 11:44 AM, Harald Schilly > wrote: > >> I'm not an expert with configuring a cluster, but I've read good things > >> about "Ansible". Do we have any ansible "experts" among us? I think, > this > >> tool is a good fit because it describes "goals" and it's extensible with > >> custom modules (written in Python). My uneducated feeling is that we > should > >> really look into it … > >> > >> -- harald > >> > >> > >> On Mon, Sep 14, 2015 at 10:13 PM, William Stein > wrote: > >>> > >>> On Mon, Sep 14, 2015 at 1:10 PM, Volker Braun > >>> wrote: > >>> > I went to a talk by the TravisCI CEO a while ago and one advice that > >>> > struck > >>> > a chord was: If you have to log into your server you are doing it > wrong. > >>> > Its > >>> > of course OK for a one-off machine to debug / play around with, but > it > >>> > just > >>> > makes administration a time sink. Its better to just erase an > instance > >>> > and > >>> > recreate it from your orchestration tool. Solves the problem of > >>> > documenting > >>> > *what* you changed as a side effect. > >>> > > >>> > On that note, why not openstack / openshift / kubernetes / ...? Of > >>> > course > >>> > you can create and manage kvm images and/or containers by hand, but > why > >>> > reinvent the wheel. > >>> > >>> Let's start fresh! I'm open to any an all ideas for how to setup the > >>> cluster, and *greatly* appreciated your suggestions. > >>> > >>> -- William > >>> > >>> > >>> > >>> > > >>> > -- > >>> > -- > >>> > You received this message because you are subscribed to the Google > >>> > Groups "sage.math users" group. > >>> > To post to this group, send email to sagemath-us...@googlegroups.com > >>> > To unsubscribe from this group, send email to > >>> > sagemath-users+unsubscr...@googlegroups.com > >>> > For more options, visit this group at > >>> > http://groups.google.com/group/sagemath-users > >>> > > >>> > --- > >>> > You received this message because you are subscribed to the Google > >>> > Groups > >>> > "sage.math users" group. > >>> > To unsubscribe from this group and stop receiving emails from it, > send > >>> > an > >>> > email to sagemath-users+unsubscr...@googlegroups.com. > >>> > > >>> > For more options, visit https://groups.google.com/d/optout. > >>> > >>> > >>> > >>> -- > >>> William (http://wstein.org) > >> > >> > >> > >> > >> -- > >> Harald Schilly -- SageMath, Inc. > >> https://cloud.sagemath.com > >> Please sign up for a $7/month SageMathCloud membership. > >> Goto Billing tab in Account -> Credit Card -> Add a subscription > > > > > > > > -- > > William (http://wstein.org) > > > > -- > William (http://wstein.org) > > -- > You received this message because you are subscribed to t
[sage-devel] Re: *.math.washington.edu hardware resources
Hi, I've created a SageMathCloud project that we can use for organization. I've added everybody in "the team" to that: https://cloud.sagemath.com/projects/f125ea34-9817-4715-b879-e26e0c7fcb7c/files/ I've also made a folder "public" https://cloud.sagemath.com/projects/f125ea34-9817-4715-b879-e26e0c7fcb7c/files/public/ which is public to the world. Anyway, this will be a good place to record schedules, have chat, layout the topology of the system, record security stuff and decisions, etc. -- William On Tue, Sep 15, 2015 at 11:55 AM, William Stein wrote: > Hi Everybody, > > I think the current Sage UW cluster team is: > >- William Stein >- Andrey Novoseltsev >- Dima Pasechnik >- Vincent Delecroix >- Thierry >- Samuel Lelievre >- Harald Schilly >- Andrew Ohana (minimal availability -- but wants to discuss) > > I think that's pretty good actually. To be in the team you have to > volunteer to help out regularly for **the next year**. In exchange > you get to help out, get admin on everything, and together we'll > decide on what to do. > > Every option is on the table to start. > > There were some proposals above already about how to set things up. > I propose that we come up with ideas and get ideas from the community, > but that we together have final say on what to actually do, since > we'll be doing all the work. We can vote/argue, but if there is major > contention or a tie, I'll decide things somehow. > > Our goal (and the legally agreed upon thing with my university) is to > support (a) sage developers, and (b) research mathematicians. > > Regarding hardware, there was no budget when I first sent this email, > but now there is a small budget due to a much-appreciated (!) donation > from a Sage developer. So, e.g., buying another full rack is an > option. > > -- William > > > On Tue, Sep 15, 2015 at 11:44 AM, Harald Schilly wrote: >> I'm not an expert with configuring a cluster, but I've read good things >> about "Ansible". Do we have any ansible "experts" among us? I think, this >> tool is a good fit because it describes "goals" and it's extensible with >> custom modules (written in Python). My uneducated feeling is that we should >> really look into it … >> >> -- harald >> >> >> On Mon, Sep 14, 2015 at 10:13 PM, William Stein wrote: >>> >>> On Mon, Sep 14, 2015 at 1:10 PM, Volker Braun >>> wrote: >>> > I went to a talk by the TravisCI CEO a while ago and one advice that >>> > struck >>> > a chord was: If you have to log into your server you are doing it wrong. >>> > Its >>> > of course OK for a one-off machine to debug / play around with, but it >>> > just >>> > makes administration a time sink. Its better to just erase an instance >>> > and >>> > recreate it from your orchestration tool. Solves the problem of >>> > documenting >>> > *what* you changed as a side effect. >>> > >>> > On that note, why not openstack / openshift / kubernetes / ...? Of >>> > course >>> > you can create and manage kvm images and/or containers by hand, but why >>> > reinvent the wheel. >>> >>> Let's start fresh! I'm open to any an all ideas for how to setup the >>> cluster, and *greatly* appreciated your suggestions. >>> >>> -- William >>> >>> >>> >>> > >>> > -- >>> > -- >>> > You received this message because you are subscribed to the Google >>> > Groups "sage.math users" group. >>> > To post to this group, send email to sagemath-us...@googlegroups.com >>> > To unsubscribe from this group, send email to >>> > sagemath-users+unsubscr...@googlegroups.com >>> > For more options, visit this group at >>> > http://groups.google.com/group/sagemath-users >>> > >>> > --- >>> > You received this message because you are subscribed to the Google >>> > Groups >>> > "sage.math users" group. >>> > To unsubscribe from this group and stop receiving emails from it, send >>> > an >>> > email to sagemath-users+unsubscr...@googlegroups.com. >>> > >>> > For more options, visit https://groups.google.com/d/optout. >>> >>> >>> >>> -- >>> William (http://wstein.org) >> >> >> >> >> -- >> Harald Schilly -- SageMath, Inc. >> https://cloud.sagemath.com >> Please sign up for a $7/month SageMathCloud membership. >> Goto Billing tab in Account -> Credit Card -> Add a subscription > > > > -- > William (http://wstein.org) -- William (http://wstein.org) -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
[sage-devel] Re: *.math.washington.edu hardware resources
Hi Everybody, I think the current Sage UW cluster team is: - William Stein - Andrey Novoseltsev - Dima Pasechnik - Vincent Delecroix - Thierry - Samuel Lelievre - Harald Schilly - Andrew Ohana (minimal availability -- but wants to discuss) I think that's pretty good actually. To be in the team you have to volunteer to help out regularly for **the next year**. In exchange you get to help out, get admin on everything, and together we'll decide on what to do. Every option is on the table to start. There were some proposals above already about how to set things up. I propose that we come up with ideas and get ideas from the community, but that we together have final say on what to actually do, since we'll be doing all the work. We can vote/argue, but if there is major contention or a tie, I'll decide things somehow. Our goal (and the legally agreed upon thing with my university) is to support (a) sage developers, and (b) research mathematicians. Regarding hardware, there was no budget when I first sent this email, but now there is a small budget due to a much-appreciated (!) donation from a Sage developer. So, e.g., buying another full rack is an option. -- William On Tue, Sep 15, 2015 at 11:44 AM, Harald Schilly wrote: > I'm not an expert with configuring a cluster, but I've read good things > about "Ansible". Do we have any ansible "experts" among us? I think, this > tool is a good fit because it describes "goals" and it's extensible with > custom modules (written in Python). My uneducated feeling is that we should > really look into it … > > -- harald > > > On Mon, Sep 14, 2015 at 10:13 PM, William Stein wrote: >> >> On Mon, Sep 14, 2015 at 1:10 PM, Volker Braun >> wrote: >> > I went to a talk by the TravisCI CEO a while ago and one advice that >> > struck >> > a chord was: If you have to log into your server you are doing it wrong. >> > Its >> > of course OK for a one-off machine to debug / play around with, but it >> > just >> > makes administration a time sink. Its better to just erase an instance >> > and >> > recreate it from your orchestration tool. Solves the problem of >> > documenting >> > *what* you changed as a side effect. >> > >> > On that note, why not openstack / openshift / kubernetes / ...? Of >> > course >> > you can create and manage kvm images and/or containers by hand, but why >> > reinvent the wheel. >> >> Let's start fresh! I'm open to any an all ideas for how to setup the >> cluster, and *greatly* appreciated your suggestions. >> >> -- William >> >> >> >> > >> > -- >> > -- >> > You received this message because you are subscribed to the Google >> > Groups "sage.math users" group. >> > To post to this group, send email to sagemath-us...@googlegroups.com >> > To unsubscribe from this group, send email to >> > sagemath-users+unsubscr...@googlegroups.com >> > For more options, visit this group at >> > http://groups.google.com/group/sagemath-users >> > >> > --- >> > You received this message because you are subscribed to the Google >> > Groups >> > "sage.math users" group. >> > To unsubscribe from this group and stop receiving emails from it, send >> > an >> > email to sagemath-users+unsubscr...@googlegroups.com. >> > >> > For more options, visit https://groups.google.com/d/optout. >> >> >> >> -- >> William (http://wstein.org) > > > > > -- > Harald Schilly -- SageMath, Inc. > https://cloud.sagemath.com > Please sign up for a $7/month SageMathCloud membership. > Goto Billing tab in Account -> Credit Card -> Add a subscription -- William (http://wstein.org) -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
Re: [sage-devel] Re: *.math.washington.edu hardware resources
2015-09-14 20:24:01 UTC+2, Thierry: Some few ideas i planned to work on for a while: > > - random testing : the current random doctests uses a single seed, so we > do > not get any more information when the doctests are run. We could > also have some truly random tests that, for example, test the > associativity in groups from random_elements. The more they are run, the > stronger Sage is. Then we could have som VM that only runs those random > tests while changing the seed each time. > > - deploy builder virtual machines for more distros/architectures as i did > for Debian/Ubuntu. We can think of something cleaner than > http://sagebuilder.metelu.net/ for example something based on vagrant. > > - explicit the release tests, and implement them as scripts, for example > (some release ideas come from a discussion with Jeroen): > - Sage must build and pass tests on a particular set of configurations > - try various combinations of optional packages > - compile Sage with some user and run it from some other user > - check the permissions of the files > - check if the source code has some bare 'except:' statements > - compile the pdf documentation to discover ill-formatted LaTeX > formulas > - check if openssl is up-to-date (before each official realease) > - ... > - more generaly, write a test script for each problem that once > appeared on sage-release > Happy to help too. One extra thing we could do is have some machines (or virtual machines) running Debian, Ubuntu, Fedora, ... and testing that installing the SageMath packages for those distros (PPA and similar) works. We could launch them and interactively test a few basic things such as sage: 1+1 sage: plot(x^2) sage: plot3d(lambda x,y,z: x^2+y^2+z^2) sage: range? and also run automated long tests. Some links related to SageMath in various Linux distros: arch-linux: sagemath package, maintained by Antonio Rojas https://www.archlinux.org/packages/community/x86_64/sagemath/ Fedora: sagemath package, owned by Paulo César Pereira de Andrade (pcpa) https://apps.fedoraproject.org/packages/sagemath https://fedoraproject.org/wiki/SIGs/SciTech/SAGE RPM package manager: sagemath package http://rpmfind.net/linux/rpm2html/search.php?query=sagemath Ubuntu: sagemath ppa, by AIMS-team https://launchpad.net/~aims/+archive/ubuntu/sagemath Debian: might be worth making a .deb package for SageMath, as this comes up every so often on ask-sage, sage-support, sage-devel. -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
Re: [sage-devel] Re: *.math.washington.edu hardware resources
On 09/14/2015 02:23 PM, Thierry wrote: > Hi, > > i am pretty remote right now (for a Sage tutorial actually;), but you can > count me in, starting from the next week. > > Some few ideas i planned to work on for a while: > > - random testing : the current random doctests uses a single seed, so we do > not get any more information when the doctests are run. We could > also have some truly random tests that, for example, test the > associativity in groups from random_elements. The more they are run, the > stronger Sage is. Then we could have som VM that only runs those random > tests while changing the seed each time. > I've been meaning to bring this up. We need to get rid of the fixed seed. I know for sure I've written random tests in the past, not knowing that sage silently replaces its random number generator with https://xkcd.com/221/. Unless you happen to stumble upon the "Library API" section of http://doc.sagemath.org/html/en/reference/misc/sage/misc/randstate.html the fact that your random numbers have been subverted will not be apparent unless you find some bug in your code that should have been doctested. The rationale given is that we need to be able to reproduce failing doctests. Well, why not just spit out the random seed at the beginning of `sage -t` then? See `random_cone()` for an example where I call set_random_seed() a million times. Had I not found a few bugs in the code that should have been tested, I probably would not have noticed that randomness was broken, and the function might have been merged with bugs. -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
[sage-devel] Re: *.math.washington.edu hardware resources
On Mon, Sep 14, 2015 at 1:10 PM, Volker Braun wrote: > I went to a talk by the TravisCI CEO a while ago and one advice that struck > a chord was: If you have to log into your server you are doing it wrong. Its > of course OK for a one-off machine to debug / play around with, but it just > makes administration a time sink. Its better to just erase an instance and > recreate it from your orchestration tool. Solves the problem of documenting > *what* you changed as a side effect. > > On that note, why not openstack / openshift / kubernetes / ...? Of course > you can create and manage kvm images and/or containers by hand, but why > reinvent the wheel. Let's start fresh! I'm open to any an all ideas for how to setup the cluster, and *greatly* appreciated your suggestions. -- William > > -- > -- > You received this message because you are subscribed to the Google > Groups "sage.math users" group. > To post to this group, send email to sagemath-us...@googlegroups.com > To unsubscribe from this group, send email to > sagemath-users+unsubscr...@googlegroups.com > For more options, visit this group at > http://groups.google.com/group/sagemath-users > > --- > You received this message because you are subscribed to the Google Groups > "sage.math users" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to sagemath-users+unsubscr...@googlegroups.com. > > For more options, visit https://groups.google.com/d/optout. -- William (http://wstein.org) -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
[sage-devel] Re: *.math.washington.edu hardware resources
I went to a talk by the TravisCI CEO a while ago and one advice that struck a chord was: If you have to log into your server you are doing it wrong. Its of course OK for a one-off machine to debug / play around with, but it just makes administration a time sink. Its better to just erase an instance and recreate it from your orchestration tool. Solves the problem of documenting *what* you changed as a side effect. On that note, why not openstack / openshift / kubernetes / ...? Of course you can create and manage kvm images and/or containers by hand, but why reinvent the wheel. -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
Re: [sage-devel] Re: *.math.washington.edu hardware resources
As William mentioned, I don't really have the time to work on administering the cluster, but I would be happy to part of the initial discussions (e.g what I was planning on doing if I had the time, why I chose centos, in-depth inventory of hardware, etc). -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
Re: [sage-devel] Re: *.math.washington.edu hardware resources
Hi, i am pretty remote right now (for a Sage tutorial actually;), but you can count me in, starting from the next week. Some few ideas i planned to work on for a while: - random testing : the current random doctests uses a single seed, so we do not get any more information when the doctests are run. We could also have some truly random tests that, for example, test the associativity in groups from random_elements. The more they are run, the stronger Sage is. Then we could have som VM that only runs those random tests while changing the seed each time. - deploy builder virtual machines for more distros/architectures as i did for Debian/Ubuntu. We can think of something cleaner than http://sagebuilder.metelu.net/ for example something based on vagrant. - explicit the release tests, and implement them as scripts, for example (some release ideas come from a discussion with Jeroen): - Sage must build and pass tests on a particular set of configurations - try various combinations of optional packages - compile Sage with some user and run it from some other user - check the permissions of the files - check if the source code has some bare 'except:' statements - compile the pdf documentation to discover ill-formatted LaTeX formulas - check if openssl is up-to-date (before each official realease) - ... - more generaly, write a test script for each problem that once appeared on sage-release Ciao, Thierry On Mon, Sep 14, 2015 at 02:54:19PM -0300, Vincent Delecroix wrote: > On 14/09/15 14:47, William Stein wrote: > >Who else? Once we have a small team, we'll (publicly) figure out a plan. > > You can count on me as well. > > Vincent > > -- > You received this message because you are subscribed to the Google Groups > "sage-devel" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to sage-devel+unsubscr...@googlegroups.com. > To post to this group, send email to sage-devel@googlegroups.com. > Visit this group at http://groups.google.com/group/sage-devel. > For more options, visit https://groups.google.com/d/optout. -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
[sage-devel] Re: *.math.washington.edu hardware resources
Hello, on the list of things hosted, you should add the patchbot server, which is now running in a virtual machine on CentOs. Frédéric Le lundi 14 septembre 2015 02:55:49 UTC+2, William a écrit : > > On Sun, Sep 13, 2015 at 12:44 PM, Dima Pasechnik > wrote: > > On 13 September 2015 at 11:00, William Stein > wrote: > >> On Sun, Sep 13, 2015 at 11:50 AM, Harald Schilly > >> > wrote: > >>> Well, I don't know who administers this "sage.math" server and AFAIK I > >>> don't even have root access. > >> > >> I can't login either. Andrew must be the only one with access. > >> > >> I think we should get a few people together and reset everything > >> regarding the sagemath hardware -- who wants to help? > >> > > > > Could I help somehow? > > YES. Please, please, I hope that around 3 or 4 people respond and > say, "I can definitely help". We will then -- as a community -- come > up with a plan for how to use the extensive hardware I have at UW that > fits what the internal audit office agreed with, and then do it. I > can do what needs local access, and everybody can do all the other > stuff, after we discuss it together, and come up with an approach that > has certain properties that make sense. > > We can discuss this on the sagemath-admins list in the long run. We > could start with an inventory of hardware and space, then go from > there. > > In the short run, let's think about a plan that works. My memory is > that this is the inventory of resources: > > - one single rack > > - "infinite" free bandwidth and power > > - 20 (?) Dell R415 1U servers with 16 cores and between 64 and 96GB > RAM and between 1 and 6TB of disk. Most have a 1TB SSD. > > - 4 2U Suns from Jan 2009 with 24 cores and 128GB RAM. > > - a 64-core Dell with 192GB RAM that regularly crashes. > > - 1 4U Sun with 24TB of disk (48 512GB disks) > > - a big UPS that takes up nearly 1/3 of the rack. > > I think that is it. Note that currently half of the Dell R415's are > in pile on the floor, since they don't fit in the rack. > > Andrew Ohana put CentOS and some kvm VM's on some of the above > machines. However, he has no time for admin work, so I'm OK with > starting over. > > Ideas from people doing the work to implement them *and* maintain them > for at least a year are welcome. > > I would like to have a core team of about 5 people who all have full > admin access to all machines, and understand *exactly* what is where > involved. Most of the above machines even have remote management > interfaces, which provide full console access (and CD drives, etc.). > > What we host: > >- trac >- sage wiki >- downloads/rsync server >- cython's stuff: on a dedicated server >- mpir's stuff: on a dedicated server >- a few hundred sage.math accounts and corresponding files... like > Simon King's. > > These are all *fine* to continue hosting, including Simon's stuff and > much more like it... if the community will help. > > -- William > > > > > > > > -- > William (http://wstein.org) > -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
Re: [sage-devel] Re: *.math.washington.edu hardware resources
On 14/09/15 14:47, William Stein wrote: Who else? Once we have a small team, we'll (publicly) figure out a plan. You can count on me as well. Vincent -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
[sage-devel] Re: *.math.washington.edu hardware resources
On Mon, Sep 14, 2015 at 10:37 AM, Andrey Novoseltsev wrote: > CentOS, but presumably it is not completely different from > Debian/Ubuntu). Changing everything, e.g., starting from scratch again with Ubuntu is completely on the table. We do not have to stick with CentOS. Once we get going, we will discuss what to do. A summary of the UW internal audit conditions: any research mathematician or Sage developer may use the resources for research mathematics or Sage development.Our goal would be to best use the aforementioned computers for these purposes. > I'll be happy to get involved as well (don't have much experience with Cool. So now it is: - William Stein - Andrey Novoseltsev - Dima Pasechnik Who else? Once we have a small team, we'll (publicly) figure out a plan. William > > On Mon, Sep 14, 2015 at 1:05 AM, William Stein wrote: >> On Sunday, September 13, 2015, Dima Pasechnik wrote: >>> Do you mean it's a real "out-of-band" management, a.k.a. "lights-out" >>> management >>> system? (I must say I have no idea about these). >>> If yes, what would one need on the remote end, any special hardware? >> >> >> Yes completely out of band. > > Regarding what is needed: nothing special. I've used such a system > which runs on a separate Ethernet cable with its own IP address - > going there with a browser allows monitoring and control including > power on/off, as well as remote console and CD emulation. The bad > thing was that remote console was a Java applet and due to security > settings it is tricky to run nowadays. Also, such access was provided > only for computers in the same network (sensible security measure). To > use it really remotely I had to make multiple SSH tunnels (depending > on what I had to do - different ports were used for > browser/console/emulation). Otherwise it is quite smooth and > straightforward, I had display/keyboard hooked up to this machine only > once two years ago - to turn this remote access on. > > Best, > Andrey > > -- > -- > You received this message because you are subscribed to the Google > Groups "sage.math users" group. > To post to this group, send email to sagemath-us...@googlegroups.com > To unsubscribe from this group, send email to > sagemath-users+unsubscr...@googlegroups.com > For more options, visit this group at > http://groups.google.com/group/sagemath-users > > --- > You received this message because you are subscribed to the Google Groups > "sage.math users" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to sagemath-users+unsubscr...@googlegroups.com. > For more options, visit https://groups.google.com/d/optout. -- William (http://wstein.org) -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
