Thanks! You are silently watched :-), and your effort is highly appreciated, quite for sure not just by me.
Kind regards, Steffen Barbie <bar...@missbarbell.co.uk> writes: > Everything is more or less back on track and/or catching up. Only the mailers > are left > to do for the cronjobs. I still need to test those, but all being well they > should be > running after tomorrow. SSL will hopefully be sorted at the weekend or early > next week. > > Expect things to be a bit up in the air for the rest of week, and I'll post > as and when > there are any significant updates. > > Barbie. > > -- > Birmingham.pm - http://birmingham.pm.org > CPAN Testers - http://cpantesters.org > YAPC Surveys - http://yapc-surveys.org > Perl Jam - http://perljam.info > > On Wed, Oct 2, 2013 at 8:26 AM, Barbie <bar...@missbarbell.co.uk> wrote: > > DNS has now been switched to the new server for Reports & Statistics > site. I've also > switched the Preferences site, however the SSL Cert hasn't been set up > for that yet, > so please bear with me, and I'll get those set up as soon as I can. > > Thanks, > Barbie. > > -- > Birmingham.pm - http://birmingham.pm.org > CPAN Testers - http://cpantesters.org > YAPC Surveys - http://yapc-surveys.org > Perl Jam - http://perljam.info > > On Wed, Oct 2, 2013 at 8:06 AM, Barbie <bar...@missbarbell.co.uk> wrote: > > We hit problems with the newer Metabase code. It seems the latest > version of > Metabase has radically different internals, and isn't pulling back > the reports > as expected. However, it doesn't help that Amazon's SimpleDB has the > most > useless ordering mechanism I've ever had the misfortune to deal with. > Apparently > the following timestamps are in ascending order: > > GUID [3140779c-26ba-11e3-b8b2-96b61dda783b].. time > [2013-09-26T14:44:48Z] > [2013-09-26T14:44:48Z].. stored.. cached > GUID [cc212043-6e0e-1014-82b0-0d755be8bcae].. time > [2013-09-27T13:37:20Z] > [2013-09-27T13:37:20Z].. stored.. cached > GUID [320ef5ae-26ba-11e3-b8b2-96b61dda783b].. time > [2013-09-26T14:44:49Z] > [2013-09-26T14:44:49Z].. stored.. cached > GUID [2fd58d66-26ba-11e3-b8b2-96b61dda783b].. time > [2013-09-26T14:44:45Z] > [2013-09-26T14:44:45Z].. stored.. cached > GUID [ddc56792-2779-11e3-8a37-e2882d12fa71].. time > [2013-09-27T13:36:51Z] > [2013-09-27T13:36:51Z].. stored.. cached > GUID [336112a2-26ba-11e3-b8b2-96b61dda783b].. time > [2013-09-26T14:44:51Z] > [2013-09-26T14:44:51Z].. stored.. cached > GUID [3fe51316-26ba-11e3-b8b2-96b61dda783b].. time > [2013-09-26T14:45:12Z] > [2013-09-26T14:45:12Z].. stored.. cached > GUID [35d86012-26ba-11e3-b8b2-96b61dda783b].. time > [2013-09-26T14:44:56Z] > [2013-09-26T14:44:56Z].. stored.. cached > GUID [38255816-26ba-11e3-b8b2-96b61dda783b].. time > [2013-09-26T14:44:59Z] > [2013-09-26T14:44:59Z].. stored.. cached > GUID [3d2ccc72-26ba-11e3-b8b2-96b61dda783b].. time > [2013-09-26T14:45:08Z] > [2013-09-26T14:45:08Z].. stored.. cached > GUID [45137f76-26ba-11e3-b8b2-96b61dda783b].. time > [2013-09-26T14:45:21Z] > [2013-09-26T14:45:21Z].. stored.. cached > GUID [de87e77c-2779-11e3-b8b2-96b61dda783b].. time > [2013-09-27T13:36:52Z] > [2013-09-27T13:36:52Z].. stored.. cached > > We've been hit by this numerous times, and it is why there is so much > code > inside the Generator to try and cope with this. I look forward to > when we can > finally move to MongoDB :) > > Nearly all the cronjobs are switched back on, and I hope to have the > Reports and > Statistics websites running off the new server sometime today. > > Thanks, > Barbie. > > -- > Birmingham.pm - http://birmingham.pm.org > CPAN Testers - http://cpantesters.org > YAPC Surveys - http://yapc-surveys.org > Perl Jam - http://perljam.info > > On Tue, Oct 1, 2013 at 8:06 AM, Barbie <bar...@missbarbell.co.uk> > wrote: > > After a bit of a hiccup this morning, it appears the pull process > from the > Metabase is now working. I'm going to let it run unhindered this > morning to > give it a good run at catching up. It's currently 4 days behind, > but its > shooting through at the moment, so I expect it to be up to date > in a few > days. I'll switch the builders on this afternoon or this evening, > and > hopefully I can switch the websites over tonight too. > > Cheers, > Barbie. > > -- > Birmingham.pm - http://birmingham.pm.org > CPAN Testers - http://cpantesters.org > YAPC Surveys - http://yapc-surveys.org > Perl Jam - http://perljam.info > > On Mon, Sep 30, 2013 at 8:58 PM, Barbie > <bar...@missbarbell.co.uk> wrote: > > Okay the field list on the cpanstats db completed, so I > killed the > repair, and recopied the metabase table. Now waiting for the > field list > to complete on the metabase db. Seems its not supposed to > lock, but > there is a bug in the current releases of MySQL where it does > :( > > The reports pull is now queued, so as soon as the field list > completes > it'll be off and running. I'll be switching other cronjobs on > once it > starts processing. > > Cheers, > Barbie. > > -- > Birmingham.pm - http://birmingham.pm.org > CPAN Testers - http://cpantesters.org > YAPC Surveys - http://yapc-surveys.org > Perl Jam - http://perljam.info > > On Mon, Sep 30, 2013 at 2:30 PM, Barbie > <bar...@missbarbell.co.uk> > wrote: > > Not sure of the best way forward, so if anyone has > suggestions, > please shout. > > The index broke because I started a mysql shell, without > the -A > option. As a consequence it started a process to get the > field list > of the tables in the metabase db. As it was taking over > an hour I > CTRL-C'ed it and killed the process. Bad idea. I've now > been running > the repair table for over 14 hours, as well as running a > field list > query on the cpanstats db that was started by another > mysql shell. > > The question is, is there a quick way to tell the mysql > shell what > the field list is? Or is it not getting a field list but > do > something else, as the state says 'checking table'. I'm > loathed to > stop anything at the moment, as I don't know how long it > will take, > and whether it is near the end anyway. With the split of > the disks, > I'm confident the reports pull and builder scripts will > catch up > very quickly, so I'm willing to wait another day or two. > Failing > that I'm going to have to write an alias for the command > to also add > the -A. > > Does this field list actually lock any tables, as it > appeared to > when I first started it, but some of the cronjobs I've > restarted > appear to be fine. If it doesn't lock tables, I'll stop > the repair, > recopy the backup metabase db, then try another mysql > shell, and > just let it run. > > All advice gratefully received. > > Thanks, > Barbie. > > -- > Birmingham.pm - http://birmingham.pm.org > CPAN Testers - http://cpantesters.org > YAPC Surveys - http://yapc-surveys.org > Perl Jam - http://perljam.info > > On Sun, Sep 29, 2013 at 10:05 PM, Barbie > <bar...@missbarbell.co.uk> > wrote: > > I knew it was fateful to say it had gone well. I > accidentally > managed to break the indexing during a field list > check :( > Repairing table now, and also re-copying a fresh > copy, just in > case that would be quicker. > > More progress tomorrow. > > Barbie. > > -- > Birmingham.pm - http://birmingham.pm.org > CPAN Testers - http://cpantesters.org > YAPC Surveys - http://yapc-surveys.org > Perl Jam - http://perljam.info > > On Sun, Sep 29, 2013 at 6:54 PM, Barbie < > bar...@missbarbell.co.uk> wrote: > > Databases copied. Libraries (Perl & C) installed. > Websites > prepped. Now starting to switch cronjobs back on. > > Might have some hiccups getting all the mailers > working, as > I'm sure I had problems with them last time with > reverse > DNS. > > Prefs site may need to wait until later in the > week, as I > need to sort out some new SSL certificates, both > for the > prefs.cpantesters.org and admin.cpantesters.org. > The current > one is due to expire shortly, and would rather > use someone > else than have to go through the hoops with our > current > supplier. > > Thankfully so far all has gone well. However, > please let me > know if you see any issues arising during the > week. > > Thanks, > Barbie. > > -- > Birmingham.pm - http://birmingham.pm.org > CPAN Testers - http://cpantesters.org > YAPC Surveys - http://yapc-surveys.org > Perl Jam - http://perljam.info > > On Sun, Sep 29, 2013 at 1:03 AM, Barbie < > bar...@missbarbell.co.uk> wrote: > > ftp, rsync and web now running for: > > cpan.cpantesters.org > backpan.cpantesters.org > > Databases still copying and not due to finish > until the > morning. > > Thanks, > Barbie. > > -- > Birmingham.pm - http://birmingham.pm.org > CPAN Testers - http://cpantesters.org > YAPC Surveys - http://yapc-surveys.org > Perl Jam - http://perljam.info > > On Sat, Sep 28, 2013 at 9:21 AM, Barbie < > bar...@missbarbell.co.uk> wrote: > > Minor set back, the CPAN and BACKPAN > mirrors were > too big to fit on the SSDs with the web > pages, so > having to copy them over to the HDDs. > Builders are > up to date, only page requests getting > added, rather > than reports, so they're ready to go. > Once mirrors > have been copied and configured for the > web/ftp/ > rsync, the sync from master will be > restarted. Then > the database will be next. > > Thanks, > Barbie. > > -- > Birmingham.pm - http://birmingham.pm.org > CPAN Testers - http://cpantesters.org > YAPC Surveys - http://yapc-surveys.org > Perl Jam - http://perljam.info > > On Fri, Sep 27, 2013 at 8:11 AM, Barbie < > bar...@missbarbell.co.uk> wrote: > > Just to let you know the reports pull > has now > stopped, with the last report pulled > from 06:11 > UTC. The builder will now catch up, > and as soon > as the DB is as complete as possible, > that will > be moved over. I'm trying to avoid > downtime, but > you won't see any new reports on the > site until > the processes are all switched on, on > the new > server. > > Likewise anyone trying to get > database updates > via the APIs or archived data will > only get > updates up to today. The same applies > to the > CPAN/BACKPAN mirrors, which are being > copied > over at the moment, and I aim to get > these up > and running on the new server first. > Fingers > crossed this could be this evening or > tomorrow > morning. > > I'll keep you updated of progress and > let you > know when processes are switched back > on. > > Thanks, > Barbie. > > -- > Birmingham.pm - > http://birmingham.pm.org > CPAN Testers - http://cpantesters.org > YAPC Surveys - http://yapc-surveys.org > Perl Jam - http://perljam.info > > On Wed, Sep 25, 2013 at 7:01 PM, > Barbie < > bar...@missbarbell.co.uk> wrote: > > Hi folks, > > sorry for the sort notice, but the > server > upgrade will be happening from > Friday, and will > continue over the weekend. I hadn't > quite > expected Bytemark to be so quick > getting > everything ready for us :) > > Further details on the blog - http:// > blog.cpantesters.org/diary/167 > > Please pass on to anyone you know who > might be > affected. > > Cheers, > Barbie. > > -- > Birmingham.pm - > http://birmingham.pm.org > CPAN Testers - http://cpantesters.org > YAPC Surveys - http://yapc-surveys.org > Perl Jam - http://perljam.info > -- Steffen Schwigon <s...@renormalist.net>
