May 2011 - clo-devel - mailman3.common-lisp.net

[clo-devel] Cron job to check available disk vs nagios?
by Erik Huelsmann 11 Jun '11

11 Jun '11

Hi, While working on turning back on all the cron jobs that we are running on cl-net, I found there's a cron job to check available disk space. Shouldn't we be using other software, like nagios, to monitor the services? Maybe we can write plugins/task scripts to do any of the cl-net specific tasks, but having a framework like that sounds better. Bye, Erik.

7 15

[clo-devel] Migration status: can we take down old common-lisp.net (aka elephant.common-lisp.net)?
by Erik Huelsmann 31 May '11

31 May '11

It seems Franz changed the DNS for paste.lisp.org, the last service I think which was still running on old common-lisp.net. So, is there any reason not to stop that VM now? There are probably some issues to work through before we can call new common-lisp.net fully operational though. One of the things I know of is the fact that viewCVS isn't working yet. I'll be working on that coming weekend. Another one is that the Subversion commit mails don't seem to get through. Please anybody who has any remaining issues, file bugs at admin(a)common-lisp.net so we can address in an orderly manner. Thanks in advance! Bye, Erik.

1 0

[clo-devel] Migration status [2011-05-21]
by Erik Huelsmann 23 May '11

23 May '11

common-lisp.net migration is progressing nicely: - Today I was able to migrate lisppaste to new common-lisp.net. - At the moment, I'm working to restore the functionality of many of the cron jobs. - The ticket tracker (available at admin at cl-net) is working correctly (did I mention that already?) We're serving multiple domains from common-lisp.net now and most of the services. There are some issues to be solved/ steps to be completed: * cron jobs need to be enabled [working on it] * viewvc needs to be configured correctly * DNS for paste.lisp.org should point to our new IP, or become a CNAME for paste.common-lisp.net [asked Andreas Fuchs to contact Franz system admins, Franz hosts the DNS] Questions: * have git-daemon and gitweb been configured correctly? Remarks: * Some projects stored their binaries in directories related to the old ftp service, which have been removed, because the ftp service has been disabled for years now [project admins should probably be notified] Anybody who wants / is able to comment or complement? Bye, Erik.

2 1

[clo-devel] is viewcvs down?
by Anton Vodonosov 21 May '11

21 May '11

Hello. Is viewcvs down currently, or is it moved to some subdomain? Best regards, - Anton

2 2

Re: [clo-devel] Недоставленное сообщение
by Hans Hübner 18 May '11

18 May '11

Anton, I copied the spamassassin configuration to the new box moments ago. Please let us know if your problem persists. -Hans 2011/5/18 Anton Vodonosov <avodonosov(a)yandex.ru>: > Hello, > > I am getting bounced by common-lisp.net mail server. > > Hans, I remember there was a similar problem earlier, > and you have added me to the spamassasin white list. > Probably the white list is lost during the migration. Could you fix it? > > Best regards, > - Anton > > -------- Forwarded message -------- > 18.05.2011, 01:37, mailer-daemon(a)yandex.ru: > > ********** > > Это письмо отправлено почтовым сервером yandex.ru > > К сожалению, мы вынуждены сообщить Вам о том, что Ваше письмо не может > быть отправлено одному или нескольким адресатам. Причины указаны ниже. > > Пожалуйста, не отвечайте на это сообщение. > > ********** > > This is the mail system at host yandex.ru > > I'm sorry to have to inform you that your message could not > be delivered to one or more recipients. It's attached below. > > Please, do not reply to this message. > > <cl-plus-ssl-devel(a)common-lisp.net>;: host a.mx.common-lisp.net[74.115.254.14] > said: 550 This message scored 2.3 spam points. (in reply to end of DATA > command) > > Reporting-MTA: dns; forward9.mail.yandex.net > X-Yandex-Queue-ID: 5CC7BCE180B > X-Yandex-Sender: rfc822; avodonosov(a)yandex.ru > Arrival-Date: Wed, 18 May 2011 01:26:24 +0400 (MSD) > > Final-Recipient: rfc822; cl-plus-ssl-devel(a)common-lisp.net > Original-Recipient: rfc822;cl-plus-ssl-devel(a)common-lisp.net > Action: failed > Status: 5.0.0 > Remote-MTA: dns; a.mx.common-lisp.net > Diagnostic-Code: smtp; 550 This message scored 2.3 spam points. > -------- End of the forwarded message -------- > Final-Recipient: rfc822; cl-plus-ssl-devel(a)common-lisp.net > Original-Recipient: rfc822;cl-plus-ssl-devel(a)common-lisp.net > Action: failed > Status: 5.0.0 > Remote-MTA: dns; a.mx.common-lisp.net > Diagnostic-Code: smtp; 550 This message scored 2.3 spam points. > > > ---------- Forwarded message ---------- > From: Anton Vodonosov <avodonosov(a)yandex.ru> > To: Attila Lendvai <attila.lendvai(a)gmail.com> > Date: Wed, 18 May 2011 01:26:23 +0400 > Subject: Re: [cl-plus-ssl-devel] concurrency problems when used with chrome? > Hello Attila, > > There is an outstanding multi-threading problem which so far was confirmed > for CCL + Windows. We thought for Linux it works OK (it was tested with ApacheBench). > You may find more details in the recent cl-plus-ssl-devel archives. > In short, this problem is not easy to catch, and there is a hight probability if > will remain unfixed in the nearest months. > > What is your platform (OS and Lisp impl)? Any other details you think relevant? > > As a workaround, until the problem is fixed, you may need to use a reverse proxy like > nginx for https. > > Best regards, > - Anton > > 18.05.2011, 00:49, "Attila Lendvai" <attila.lendvai(a)gmail.com>: >> dear list, >> >> i've successfully added https support to hu.dwim.web-server, which >> thanks to cl+ssl was much less trouble (for now) than i've expected. >> >> it seems to work mostly fine with ff and opera, but with chrome i get >> the following error randomly in half of the worker threads (one worker >> for one connection): >> >> The TLS/SSL connection on handle #.(SB-SYS:INT-SAP #X00C1C4A0) has >> been closed. (return code: 5)SSL error queue: >> >> [Condition of type CL+SSL::SSL-ERROR-ZERO-RETURN] >> >> the relevant part of the backtrace: >> >> 5: ((LABELS HU.DWIM.UTIL::HANDLE-LEVEL-1-ERROR) >> #<CL+SSL::SSL-ERROR-ZERO-RETURN {10088187B1}>) >> 6: (SIGNAL #<CL+SSL::SSL-ERROR-ZERO-RETURN {10088187B1}>) >> 7: (ERROR CL+SSL::SSL-ERROR-ZERO-RETURN :HANDLE #.(SB-SYS:INT-SAP >> #X7FFFEC0654B0) :RET 5 :QUEUE "SSL error queue: \n") >> 8: (CL+SSL::SSL-SIGNAL-ERROR #.(SB-SYS:INT-SAP #X7FFFEC0654B0) >> #<FUNCTION CL+SSL::SSL-READ> 5 0) >> 9: ((FLET #:WPO[READ-HTTP-REQUEST/HEAD]46)) >> 10: (HU.DWIM.WEB-SERVER::READ-HTTP-REQUEST/HEAD 15 >> #<CL+SSL::SSL-SERVER-STREAM for 15> #.(SB-SYS:INT-SAP #X7FFFEC0654B0) >> :LENGTH-LIMIT 65536) >> >> i suspect that there's something inside cl+ssl that is not threadsafe, >> because afaik chrome is more aggressively parallel than the others... >> maybe there's a C struct that should be thread local and is a mere >> global? >> >> unfortunately my first glance didn't find anything suspicious. >> >> any ideas how to proceed with this? >> >> -- >> attila >> >> Notice your eroding (digital) freedom, and do something about it! >> >> PGP: 2FA1 A9DC 9C1E BA25 A59C 963F 5D5F 45C7 DFCD 0A39 >> OTR XMPP: 8647EEAC EA30FEEF E1B55146 573E52EE 21B1FF06 >> BitCoin: 154uf86Vd9rpjMULd9CXa7nVwikknYZJiB >> >> _______________________________________________ >> cl-plus-ssl-devel mailing list >> cl-plus-ssl-devel(a)common-lisp.net >> http://lists.common-lisp.net/cgi-bin/mailman/listinfo/cl-plus-ssl-devel > >

2 1

[clo-devel] Slime docs missing?
by Zach Beane 17 May '11

17 May '11

http://common-lisp.net/projects/slime/ has many broken images on it and the link to the manual, http://common-lisp.net/projects/slime/doc/html/, is broken - it gives a 404. What happened? Is there anything I can do to help fix? Zach

2 1

[clo-devel] SLIME updates warning
by Zach Beane 16 May '11

16 May '11

I get this message when I try to update SLIME from anonymous CVS: cvs update: warning: cannot open history file `/project/slime/cvsroot/CVSROOT/history' for write: Permission denied What should I do? Zach

2 2

[clo-devel] Migration status update [2011-05-13]
by Erik Huelsmann 14 May '11

14 May '11

As Hans wrote in a mail just ago, we're progressing nicely on the migration to the new server. These steps have been completed: * MTA is now on the new server * lists.common-lisp.net is served by the new server [traffic to http://common-lisp.net/{mailman/,pipermail/} is forwarded to http://lists.common-lisp.net/...] * trac.common-lisp.net is served by the new server * ssh traffic is handled by the new server (except for just a few of us who still need access to the old server) * the ticketing system is now on the new server (old tickets will be handled from the old system; data is not migrated) * the cron jobs are gradually being turned on [only partially done] Remaining steps: * configure the remaining cron jobs * configure the common-lisp.net vhost (with the same forwarding rules that are on old-* to forward the mailman links) * configure viewcvs * compile an SBCL which will run the IRC bots stably (Stelian works on it) I don't think we need WebSVN anymore: we have Trac now. Any other steps I forgot? (Probably lots...) Bye, Erik.

2 1

[clo-devel] New rt installation
by Hans Hübner 13 May '11

13 May '11

Hi, I have just installed a fresh copy of RT for request tracking on the new box. I prefer rt to otrs because it allows me to handle requests mostly by email, which is something that otrs does not seem to do. Currently, Erik and myself are on the list of recipients for new tickets. We'd be happy to see more folks join in. If you are interested in getting an rt account and rights to handle incoming emails, please let me know and I'll set things up for you. -Hans

1 0

[clo-devel] Abandoned projects
by Hans Hübner 13 May '11

13 May '11

Hi, what do we want to do with abandoned projects? By abandoned, I mean projects where the owner has become unreachable at the email address that he has provided or configured in their .forward file? Several users of common-lisp.net are no longer reachable by email, and mailman bounces moderation reminder emails back to us, which does not make a whole lot of sense. I tend to think that deleting (archiving) such project mailing lists would be best. If anyone is interested, they can always get in touch with us. Thoughts? Hans

5 8

[clo-devel] Migration status [2011-05-06]
by Erik Huelsmann 12 May '11

12 May '11

These things happened last week: * ssh connections to cl-net were routed to new-cl-net * we were asked to install several applications on new-cl-net (emacs, zsh) * we started to create subdomains for services (first: git.cl.net) * after being reported that people were missing group memberships, we've now migrated all project related groups with a gid < 1000; because those numbers are reserved for system groups, they were excluded from the account/group migration. In order to prevent similar issues in the future, all these groups have been renumbered. There's currently a batch running to re-assign all group ownerships in the file systems * we stopped routing traffic to new-cl-net because commit mails were no longer coming through [the mailer wasn't fully set up yet on new-cl-net] Next steps: * set up the mailer * re-instate the ssh traffic route to new-cl-net * set up the webserver in order to be able to migrate trac.c-l.net Anybody who wants to use new-cl-net services directly should log into tiger.common-lisp.net; passwords and ssh keys are the same as on common-lisp.net (aka elephant.common-lisp.net) Bye, Erik.

3 6

Re: [clo-devel] Commit fail on USOCKET SVN
by Hans Hübner 11 May '11

11 May '11

problem fixed. On Wednesday, May 11, 2011, Chun Tian (binghe) <binghe.lisp(a)gmail.com> wrote: > Hi, Hans > > Still cannot commit. But I tried my other project's SVN on cl.net, i.e. cl-net-snmp, svn commit is OK. So I guess this maybe a usocket specific issue. Now I've done a full dump of usocket svn data, and try to rebuild it into a new repository directory. > > But I need some root help, the first is following command: > > chmod g+w /project/usocket > > This let me (ctian, member of usocket group) could modify project root directory to create any new top-level directory. > > Regards, > > Chun Tian (binghe) > > 在 2011-5-11，13:47， Hans Hübner 写道： > >> Drew fixed the box and I manually mounted the nfs file systems. >> Please try again! >> >> -Hans >> >> 2011/5/11 Hans Hübner <hans.huebner(a)gmail.com>: >>> Oh well, I rebooted the box, but it fails to come up again. Drew will >>> have to take care of this. Sorry. >>> >>> -Hans >>> >>> 2011/5/11 Chun Tian (binghe) <binghe.lisp(a)gmail.com>: >>>> Hi, Hans >>>> >>>> Thanks for taking care this, but I'm afraid that the problem is still there. Below is what I did in past minutes: >>>> >>>> """ >>>> binghe@binghe-i7:~/Lisp/usocket/branches/0.5.x/test$ LANG=C svn ci -m "[test] Issue elliott-slaughter.2 (WAIT-FOR-INPUT/win32 on TCP socket)" >>>> Sending test/wait-for-input.lisp >>>> Transmitting file data .svn: Commit failed (details follow): >>>> svn: Can't get exclusive lock on file '/project/usocket/svn/db/write-lock': No locks available >>>> """ >>>> >>>> As a Linux administrator, I think the theory behind your NFS fix is good, but to make the configuration change actually applied to current system, you have to completely unmount the NFS volumes on common-lisp.net, and do a full restart of NFS related services and then mount them again. Or, A simply reboot of common-lisp.net (I'm not sure if it's virtual machine) could achieve the goal easily. But any of the two choices would cause services temporary shutdown. >>>> >>>> Regards, >>>> >>>> Chun Tian (binghe) >>>> >>>> 在 2011-5-11，12:38， Hans Hübner 写道： >>>> >>>>> Hi Binghe, >>>>> >>>>> thank you for reporting the problem. I have made a change in the NFS >>>>> configuration (http://gibbonsr.net/2009/02/nfs-and-subversion-it-can-work/) >>>>> so that Subversion should now work. Can you please give it a try and >>>>> let me know if it is okay? >>>>> >>>>> common-lisp.net is currently being migrated to a new "box", which is >>>>> why there may be some issues in the coming days and week. Please >>>>> don't hesitate to report them. >>>>> >>>>> Thanks, >>>>> Hans >>>>> >>>>> 2011/5/11 Chun Tian <binghe.lisp(a)gmail.com>: >>>>>> Hi, Hans >>>>>> >>>>>> Are you still an administrator of common-lisp.net? Recently I cannot commit to USOCKET SVN any more, my SVN client said: >>>>>> >>>>>>> "Commit failed: Can't get exclusive lock on file '/project/usocket/svn/db/write-lock': No locks available" >>>>>> >>>>>> I think this may be caused by NFS on common-lisp.net, please help me confirm if any other developer met the same issue. >>>>>> >>>>>> Thanks, >>>>>> >>>>>> Chun Tian (binghe) >>>>>> >>>>>> 下面是被转发的邮件： >>>>>> >>>>>>> 发件人: "Chun Tian (binghe)" <binghe.lisp(a)gmail.com> >>>>>>> 日期: 2011年5月10日 23时52分42秒+0800 >>>>>>> 收件人: Erik Huelsmann <ehuels(a)gmail.com> >>>>>>> 主题: Commit fail on USOCKET SVN >>>>>>> >>>>>>> Hi, Erik >>>>>>> >>>>>>> I think USOCKET SVN have something wrong, because I cannot commit new code to it, the error message is: >>>>>>> >>>>>>> "Commit failed: Can't get exclusive lock on file '/project/usocket/svn/db/write-lock': No locks available" >>>>>>> >>>>>>> I

1 0

Re: [clo-devel] Commit fail on USOCKET SVN
by Hans Hübner 11 May '11

11 May '11

Drew fixed the box and I manually mounted the nfs file systems. Please try again! -Hans 2011/5/11 Hans Hübner <hans.huebner(a)gmail.com>: > Oh well, I rebooted the box, but it fails to come up again. Drew will > have to take care of this. Sorry. > > -Hans > > 2011/5/11 Chun Tian (binghe) <binghe.lisp(a)gmail.com>: >> Hi, Hans >> >> Thanks for taking care this, but I'm afraid that the problem is still there. Below is what I did in past minutes: >> >> """ >> binghe@binghe-i7:~/Lisp/usocket/branches/0.5.x/test$ LANG=C svn ci -m "[test] Issue elliott-slaughter.2 (WAIT-FOR-INPUT/win32 on TCP socket)" >> Sending test/wait-for-input.lisp >> Transmitting file data .svn: Commit failed (details follow): >> svn: Can't get exclusive lock on file '/project/usocket/svn/db/write-lock': No locks available >> """ >> >> As a Linux administrator, I think the theory behind your NFS fix is good, but to make the configuration change actually applied to current system, you have to completely unmount the NFS volumes on common-lisp.net, and do a full restart of NFS related services and then mount them again. Or, A simply reboot of common-lisp.net (I'm not sure if it's virtual machine) could achieve the goal easily. But any of the two choices would cause services temporary shutdown. >> >> Regards, >> >> Chun Tian (binghe) >> >> 在 2011-5-11，12:38， Hans Hübner 写道： >> >>> Hi Binghe, >>> >>> thank you for reporting the problem. I have made a change in the NFS >>> configuration (http://gibbonsr.net/2009/02/nfs-and-subversion-it-can-work/) >>> so that Subversion should now work. Can you please give it a try and >>> let me know if it is okay? >>> >>> common-lisp.net is currently being migrated to a new "box", which is >>> why there may be some issues in the coming days and week. Please >>> don't hesitate to report them. >>> >>> Thanks, >>> Hans >>> >>> 2011/5/11 Chun Tian <binghe.lisp(a)gmail.com>: >>>> Hi, Hans >>>> >>>> Are you still an administrator of common-lisp.net? Recently I cannot commit to USOCKET SVN any more, my SVN client said: >>>> >>>>> "Commit failed: Can't get exclusive lock on file '/project/usocket/svn/db/write-lock': No locks available" >>>> >>>> I think this may be caused by NFS on common-lisp.net, please help me confirm if any other developer met the same issue. >>>> >>>> Thanks, >>>> >>>> Chun Tian (binghe) >>>> >>>> 下面是被转发的邮件： >>>> >>>>> 发件人: "Chun Tian (binghe)" <binghe.lisp(a)gmail.com> >>>>> 日期: 2011年5月10日 23时52分42秒+0800 >>>>> 收件人: Erik Huelsmann <ehuels(a)gmail.com> >>>>> 主题: Commit fail on USOCKET SVN >>>>> >>>>> Hi, Erik >>>>> >>>>> I think USOCKET SVN have something wrong, because I cannot commit new code to it, the error message is: >>>>> >>>>> "Commit failed: Can't get exclusive lock on file '/project/usocket/svn/db/write-lock': No locks available" >>>>> >>>>> I searched on Internet [1], but all possible reasons (file system permission, disk full, svnadmin recover) have been checked, nothing found and I still cannot commit. It's a shame to me because I'm a SVN administrator in my working company ... >>>>> >>>>> Please confirm if you can commit anything to USOCKET SVN, or just help me solve this. >>>>> >>>>> Regards, >>>>> >>>>> Chun Tian (binghe) >>>>> >>>>> [1] http://www.opensubscriber.com/message/users@subversion.tigris.org/7277559.h… >>>>> >>>>> >>>> >>>> >> >> >

1 0

Re: [clo-devel] Commit fail on USOCKET SVN
by Hans Hübner 11 May '11

11 May '11

Hi Binghe, thank you for reporting the problem. I have made a change in the NFS configuration (http://gibbonsr.net/2009/02/nfs-and-subversion-it-can-work/) so that Subversion should now work. Can you please give it a try and let me know if it is okay? common-lisp.net is currently being migrated to a new "box", which is why there may be some issues in the coming days and week. Please don't hesitate to report them. Thanks, Hans 2011/5/11 Chun Tian <binghe.lisp(a)gmail.com>: > Hi, Hans > > Are you still an administrator of common-lisp.net? Recently I cannot commit to USOCKET SVN any more, my SVN client said: > >> "Commit failed: Can't get exclusive lock on file '/project/usocket/svn/db/write-lock': No locks available" > > I think this may be caused by NFS on common-lisp.net, please help me confirm if any other developer met the same issue. > > Thanks, > > Chun Tian (binghe) > > 下面是被转发的邮件： > >> 发件人: "Chun Tian (binghe)" <binghe.lisp(a)gmail.com> >> 日期: 2011年5月10日 23时52分42秒+0800 >> 收件人: Erik Huelsmann <ehuels(a)gmail.com> >> 主题: Commit fail on USOCKET SVN >> >> Hi, Erik >> >> I think USOCKET SVN have something wrong, because I cannot commit new code to it, the error message is: >> >> "Commit failed: Can't get exclusive lock on file '/project/usocket/svn/db/write-lock': No locks available" >> >> I searched on Internet [1], but all possible reasons (file system permission, disk full, svnadmin recover) have been checked, nothing found and I still cannot commit. It's a shame to me because I'm a SVN administrator in my working company ... >> >> Please confirm if you can commit anything to USOCKET SVN, or just help me solve this. >> >> Regards, >> >> Chun Tian (binghe) >> >> [1] http://www.opensubscriber.com/message/users@subversion.tigris.org/7277559.h… >> >> > >

1 0

[clo-devel] Managing (version controlling) the common-lisp.net main website
by Erik Huelsmann 08 May '11

08 May '11

The clo project contains a copy of the c-l.net website in its repository. This repository is checked out in the public_html directory of the clo project and synced hourly from there to the main web pages in /var/www. However, if I look at the /var/www directory, it contains much more than what's in the clo directory and repository. To resolve this situation, I'd like to copy all files in /var/www to the clo project and add them to the repository. There were ideas from others - I believe Raymond and Stelian - to use a different solution to maintain the website; maybe using a CMS or wiki-based system which allows WYSIWYG editing on the web. For now, I'm going with my "copy back" solution. However, it looks to me like we do need *some* kind of enhanced control. (Maybe even just a description of location of different kinds of files for maintainers to consult?) Bye, Erik.

1 0

[clo-devel] cvsroot syncing regularly takes too long
by Erik Huelsmann 08 May '11

08 May '11

Hi, On new-cl-net, the "sync cvsroot" script is regularly exceeding its runtime window (which is 20 minutes). One solution could be to change the nice parameter (-20). I'm hoping that others have better ideas, like changing the script from polling to event-based, which would eliminate the need to check all files individually. Does anybody have some cycles to look into the options? Bye, Erik.

2 2

[clo-devel] Security update required for exim4
by Hans Hübner 07 May '11

07 May '11

Drew, it seems that the Debian distribution that is installed on tiger does not supply a very current version of exim4, and today a security hole was reported in the version that we are running. Is it possible to add a package source with up to date security fixes so that we can update exim4? Thanks, Hans

1 0

Re: [clo-devel] link to OTRS docs for mail processing
by Hans Hübner 07 May '11

07 May '11

Hi Erik, I have installed the otrs gateway at otrs(a)tiger.common-lisp.net - As otrs runs under a specific user ID and I did not feel like inventing a suid wrapper, I have implemented the gateway by an explicit exim configuration directive (see /etc/exim4/configure if you are interested). I sent a few test mails to that address, but otrs did not reply to them, and I don't have an account in otrs yet. Can you add that for me, please? Thanks! Hans On Sat, May 7, 2011 at 2:33 PM, Erik Huelsmann <ehuels(a)gmail.com> wrote: > Hi Hans, > > This link discusses it: http://doc.otrs.org/3.0/en/html/x2048.html > > Basically, you need to pipe the mail into 'bin/PostMaster.pl' (which > is in /usr/share/otrs :-l ) > > The docs are for OTRS 3.0 while we still run 2.4, but there are few > changes in this area. > > Bye, > > Erik. >

1 0

[clo-devel] Testing exim4+mailman+sa+greylistd setup on tiger.common-lisp.net
by Hans Hübner 07 May '11

07 May '11

When you're reading this, things should be in working order. -Hans

1 0

[clo-devel] testing mailman+exim
by root 07 May '11

07 May '11

nothing else

1 0