| 00:04.21 | *** join/#gllug gregj_ (n=gj@fcu5.internetdsl.tpnet.pl) |
| 00:25.41 | *** join/#gllug catalyst (n=catalyst@allegro/user/angelchild) |
| 00:50.05 | *** join/#gllug Leeds (n=richard@n219078046026.netvigator.com) |
| 01:17.56 | *** join/#gllug Discordian (n=ch@chills.demon.co.uk) |
| 01:18.37 | Discordian | Blargh |
| 01:30.05 | Discordian | feels a migration coming on |
| 01:30.14 | *** part/#gllug aguai (n=aguai@host86-136-85-75.range86-136.btcentralplus.com) |
| 01:34.06 | *** join/#gllug secretlondon (n=secretlo@wikipedia/secretlondon) |
| 01:36.34 | Discordian | Hi |
| 01:36.38 | Leeds | ho? |
| 03:19.42 | *** join/#gllug Discordian (n=ch@chills.demon.co.uk) |
| 03:21.01 | *** join/#gllug curson_ (n=curzon@p2224-ipbf215funabasi.chiba.ocn.ne.jp) |
| 03:22.22 | Discordian | Hi |
| 04:01.58 | *** join/#gllug curson (n=curzon@p2224-ipbf215funabasi.chiba.ocn.ne.jp) |
| 04:27.52 | *** join/#gllug Discordian (n=ch@chills.demon.co.uk) |
| 06:29.16 | AndyMillar | mornin |
| 06:47.44 | *** join/#gllug londo__ (n=georgiou@heppc218.hep.ph.ic.ac.uk) |
| 07:06.31 | *** join/#gllug curson (n=curzon@p2224-ipbf215funabasi.chiba.ocn.ne.jp) |
| 07:29.47 | Cope | ibot seen hali |
| 07:29.49 | ibot | hali is currently on #gllug (10d 4h 54m 24s). Has said a total of 210 messages. Is idling for 9h 41m 33s, last said: 'so yeah, we've been busy :)'. |
| 08:03.27 | hali | yo |
| 08:16.53 | hali | bbl |
| 08:45.30 | *** join/#gllug tink044 (i=57c2dd50@gateway/web/ajax/mibbit.com/x-036400ce14cb198b) |
| 08:45.47 | *** join/#gllug Blapto (n=martin@62.189.100.228) |
| 08:56.16 | *** join/#gllug ElPenguin (n=ElPengui@foresight/developer/elpenguin) |
| 09:53.32 | *** join/#gllug nDj_InsticT (n=marcus@87-194-124-8.bethere.co.uk) |
| 10:12.24 | *** join/#gllug Blapto (n=martin@62.189.100.228) |
| 11:03.53 | *** part/#gllug Leeds (n=richard@n219078046026.netvigator.com) |
| 11:03.56 | *** join/#gllug huw-l (n=huw@w1199.insrv.cf.ac.uk) |
| 11:18.15 | *** join/#gllug ElPengui1 (n=ElPengui@host81-149-64-80.in-addr.btopenworld.com) |
| 11:21.34 | *** join/#gllug chrismil (n=chris@87-194-125-184.bethere.co.uk) |
| 11:23.27 | *** part/#gllug chrismil (n=chris@87-194-125-184.bethere.co.uk) |
| 12:25.42 | *** join/#gllug catalyst (n=catalyst@allegro/user/angelchild) |
| 12:40.47 | *** join/#gllug dick_turpin (n=dick_tur@host217-34-163-30.in-addr.btopenworld.com) |
| 12:46.15 | *** join/#gllug jdavies (n=jon@ubuntu/member/jdavies) |
| 14:23.41 | hali | is it just me or is vmware server 2.0b2 utter shit |
| 14:23.51 | AndyMillar | not used :( |
| 14:24.00 | hali | the new vi-look-alike interface is so buggy and slow |
| 14:24.10 | AndyMillar | i stick to 1.x |
| 14:33.06 | ChoHag | Despite being a linux admin, I don't have access to vmware server. |
| 14:33.15 | ChoHag | I am a mere user on the VMware setup. |
| 14:33.55 | *** join/#gllug nDj_InsticT (n=marcus@87-194-124-8.bethere.co.uk) |
| 14:33.55 | hali | well, im gonna go bitch to someone about spending 9k on an esx license *grrr* |
| 14:34.14 | ChoHag | For doing so, or in order to do so? |
| 14:34.16 | AndyMillar | :o |
| 14:34.24 | AndyMillar | use vmware server 1.0.5 |
| 14:35.43 | hali | to spend it because i don't like normal vmware server |
| 14:35.48 | hali | should be ok |
| 15:01.31 | Blapto | We use 1.something |
| 15:01.38 | Blapto | 2 is apparently a bit hairy |
| 15:08.54 | *** join/#gllug Discordian (n=ch@chills.demon.co.uk) |
| 15:12.32 | Discordian | Re |
| 15:19.31 | Discordian | begins |
| 15:48.11 | z00dax | hali: the 2.x series is a bit naff |
| 15:48.33 | z00dax | aparently its a lot better under the hood - but getting to under the hood needs java plugins in browser and an apple sacrifice etc |
| 15:48.52 | z00dax | I had a play with it a few months back, rolled back to 1.0.5 in about an hour |
| 15:48.56 | boudiccas | an apple sacrifice, not the usual chicken sacrifice? |
| 15:49.13 | Discordian | quicktime? |
| 15:49.48 | z00dax | :D |
| 15:50.04 | ChoHag | Fruit sacrifices are not pleasing to your god. |
| 15:50.11 | ChoHag | Only livestock will suffice. |
| 15:50.22 | Discordian | Dead goats for scsi chains |
| 16:03.51 | Discordian | Oook nasty debian/ubuntu snafu |
| 16:06.38 | antiphase | It sounds like Ubuntu have made a pig's ear of the fix, certainly |
| 16:07.00 | Discordian | Yep |
| 16:07.01 | Blapto | What's this then? |
| 16:07.26 | Discordian | http://it.slashdot.org/it/08/05/13/1533212.shtml |
| 16:07.55 | Blapto | Oh |
| 16:32.27 | jdavies | regenerated a new key |
| 16:35.24 | Discordian | I can see that being an admin nightmare |
| 18:18.34 | hali | hm, i haven't used my desktop pc for so i long i forgot how good my screen was compared to my laptop and work lcd |
| 18:18.50 | hali | Cope: in da house? |
| 18:43.33 | Cope | yup |
| 18:46.18 | Cope | hali: Indeed I am |
| 18:49.20 | hali | yo |
| 18:49.26 | Cope | :) |
| 18:49.30 | hali | so, shared nothing will suck :) |
| 18:49.39 | Cope | yeah? |
| 18:49.41 | hali | i had a think about this earlier |
| 18:49.51 | hali | i'd get two fairly cheapish iscsi 1U disk arrays |
| 18:49.59 | hali | then let oracle mirror them using oracle ASM |
| 18:50.37 | hali | and use either redhat cluster server / halinux or orale clusterware (free too) to do the failover |
| 18:50.44 | Cope | is that included in SE ? |
| 18:50.47 | hali | yes |
| 18:51.02 | hali | any licensed oracle product is allowed to user cluster ware |
| 18:51.03 | hali | but |
| 18:51.10 | hali | you need to write a basic agent script yourself |
| 18:51.15 | Cope | yeah |
| 18:51.31 | Cope | oh sorry i meant asm |
| 18:51.35 | hali | yes that too |
| 18:51.37 | Cope | i know clusterware is |
| 18:51.47 | hali | asm is included in all oracle db versions |
| 18:51.55 | Cope | right |
| 18:52.00 | hali | actually, it's the only supported storage method for anything but EE |
| 18:52.44 | Cope | so in this scheme i'm using 2 servers and 2 1u iscsi arrays |
| 18:52.54 | hali | im making the assumption that disk i/o is not critical and the db is not huge |
| 18:53.02 | Cope | DB is 60G |
| 18:53.17 | hali | hm, not tiny |
| 18:53.19 | Cope | i/o isn't a problem atm on 1 x windows box |
| 18:54.21 | hali | oracle has a pdf describing a similar setup somehwere, but google isn't helping me |
| 18:54.54 | Cope | redhat had a doc too that i couldn't find |
| 18:54.57 | hali | and you probably need two dedicted switches unless you have gig ports for a spare 8'ish port vlan |
| 18:55.08 | hali | dedicated* |
| 18:55.45 | Cope | this is getting to be more hassle that a single disk array with a gigE switch and generic clunk-over |
| 18:56.17 | hali | well, you can build it with with zero single points of failure |
| 18:56.21 | hali | for fairly low money |
| 18:56.34 | Cope | yes |
| 18:57.12 | hali | and it can be migrated to RAC in about 2 hours |
| 18:57.25 | hali | and £16k :) |
| 18:58.13 | Cope | but you think drbd is only for the barking? |
| 18:58.33 | Cope | with oracle + clusterware |
| 18:58.34 | hali | i haven't used it enough |
| 18:58.39 | Cope | nods. |
| 18:58.41 | hali | it used to be shit |
| 18:58.50 | hali | but z00dax keeps saying its ok these days :) |
| 18:59.03 | Cope | i know of some companies using it with oracle in production |
| 18:59.06 | Cope | in .de |
| 18:59.06 | hali | and i've seen some pretty cool xen+drbd implementations |
| 18:59.22 | z00dax | you can put ocfs2 on drbd and get 2 way replication these days, and it does work as long as network is not a problem |
| 18:59.50 | Cope | z00dax: you know someone who's done that? |
| 19:00.07 | hali | google drbd+ocfs2+xen |
| 19:00.15 | z00dax | Cope: there should be plenty of people |
| 19:00.15 | hali | there are some good examples |
| 19:00.28 | Cope | nods. |
| 19:00.34 | hali | i researched that a while back for a two node xen cluster with no shared disk |
| 19:00.43 | hali | then i found a spare netapp and just abused that :) |
| 19:00.54 | Cope | giggles. |
| 19:01.45 | Cope | z00dax: we should do that for mysql/postgres |
| 19:02.29 | cpufreak | erm |
| 19:02.35 | cpufreak | I wouldn't use ocfs2 ever. |
| 19:03.33 | Discordian | Once I have F9 I might look at ext4 |
| 19:04.24 | z00dax | cpufreak: ? |
| 19:04.52 | cpufreak | I found it to be utterly unstable |
| 19:04.56 | cpufreak | and when it was working, it was shit slow |
| 19:05.06 | cpufreak | by unstable, i mean it was kernel panicking nodes, in supported configurations |
| 19:05.10 | z00dax | thats not good. so well, gfs2 + drbd then |
| 19:05.23 | Cope | cpufreak: gfs2 is an other option |
| 19:05.25 | cpufreak | wouldn't use gfs2 either, it doesn't scale particularly well. |
| 19:05.31 | Cope | nods at z00dax |
| 19:05.35 | cpufreak | uses GPFS |
| 19:05.52 | Cope | hasn't used GPFS; |
| 19:05.54 | z00dax | I dont think Cope has more than a few TiB in mind |
| 19:06.02 | cpufreak | its *cheap* |
| 19:06.03 | z00dax | gpfs is good too |
| 19:06.04 | cpufreak | and fantastic. |
| 19:06.20 | cpufreak | and you don't need to fuck about drbd |
| 19:06.26 | Discordian | isn't GPFS non-free? |
| 19:06.31 | cpufreak | you can just tell it to replicate |
| 19:06.34 | Cope | looks into it. |
| 19:06.49 | cpufreak | its like 1000 euro/server or something |
| 19:06.51 | cpufreak | so its as good as free |
| 19:06.58 | z00dax | Discordian: yup, non free |
| 19:07.16 | Cope | top tip, thanks dude |
| 19:07.33 | Discordian | mind you commercial Oracle is spendy too |
| 19:07.48 | cpufreak | they don't like you running it on non-Redhat or non-SuSE |
| 19:08.01 | Cope | yeah, 1000 euro a server is easily achievable |
| 19:08.11 | cpufreak | but they will support you if you ask in the right way |
| 19:08.17 | Discordian | i'd always choose Redhat or Centos |
| 19:08.33 | Discordian | I don't much like Suse |
| 19:10.20 | Cope | i'm on rhel5, so no probs there |
| 19:10.39 | z00dax | might be worth asking ibm if they support el5 as yet |
| 19:11.02 | Discordian | Though I use Fedora at home |
| 19:12.21 | Cope | z00dax: goof point; this'll be a rebuild anyway, so could do el4 if nec. |
| 19:13.17 | Discordian | I found my SCO Oracle 7 CDs yesterday |
| 19:13.35 | Discordian | Seems ages since I ran that using ibcs2 on linux |
| 19:21.59 | Discordian | fsck still 4 hours+ |
| 19:25.27 | Cope | reads about gpfs |
| 19:26.10 | Discordian | i wish ODs-2 was in the mainstream kernel |
| 19:26.25 | Cope | smiles - how lovely to be reading IBM docs again |
| 19:26.34 | Cope | misses AIX. |
| 19:26.39 | Cope | AIX is the tits. |
| 19:26.57 | Discordian | I'd like a box that could run AIX |
| 19:27.15 | Discordian | and OpenVMS |
| 19:28.21 | Cope | had a P590 |
| 19:28.24 | Cope | well 2 |
| 19:28.42 | Discordian | At home? |
| 19:28.51 | Cope | heh no |
| 19:28.59 | Cope | $former_client |
| 19:29.17 | Discordian | they're kinda big |
| 19:30.21 | Cope | you got a new job, yet? |
| 19:31.13 | Discordian | Interview tomorrow |
| 19:31.19 | Cope | woo! where? what? |
| 19:31.40 | Discordian | Openanswers in Chessington |
| 19:31.57 | Discordian | Though it'd be teleworking mostly |
| 19:32.12 | Cope | nods; groovy. |
| 19:32.31 | Discordian | s/Chessington/Hammersmith/ |
| 19:33.31 | Discordian | A week ion hospital slowed stuff down a bit |
| 19:34.49 | Cope | eww |
| 19:34.51 | Cope | you ok? |
| 19:35.13 | Discordian | Dry gangrene on a toe |
| 19:35.31 | Cope | trenchfoot |
| 19:35.41 | Cope | BOFH standard issue |
| 19:35.59 | Discordian | Which is why a mostly teleworking job would be ideal right now |
| 19:36.18 | Cope | yeah |
| 19:37.13 | Discordian | besides the fact I could run what OS I wanted at home |
| 19:37.35 | Discordian | hated being forced to run XP |
| 20:04.57 | Discordian | Oook tis my b'day this weekend |
| 20:05.07 | *** join/#gllug londo_ (n=georgiou@heppc218.hep.ph.ic.ac.uk) |
| 21:14.19 | Cope | hali: still around? |
| 21:15.26 | hali | yes |
| 21:15.38 | hali | reading NSA system hardening papers :) |
| 21:15.51 | Cope | http://www.calivia.com/oracle-db-cluster-with-linux-ha-and-ocfs2 |
| 22:09.53 | *** join/#gllug secretlondon (n=secretlo@wikipedia/secretlondon) |
| 22:14.11 | Discordian | Lo |
| 22:34.58 | Discordian | finally a decent speed for F9 |
| 22:42.28 | secretlondon | if you are using a debian derived distro you need to remake your security keys http://www.debian.org/security/2008/dsa-1571 |
| 22:42.35 | secretlondon | bah |
| 22:42.53 | Discordian | yeah big gotcha |
| 22:44.53 | antiphase | pities the fools with commercial SSL certs with Debian-generated keys |
| 22:45.14 | wethrin | Well, if they do insist on using Debian in production...... |
| 22:45.40 | Discordian | Good news for Thawte/Verisign |
| 22:51.23 | secretlondon | well Thawte used to be based on debian |
| 22:51.58 | Discordian | heh |