| 00:43.01 | *** join/#gllug DiscordianUK (n=ch@78.146.172.73) |
| 07:50.45 | bilarh | mindbendr: direct attached with fibre |
| 08:45.45 | hali | morning |
| 08:46.09 | hali | bilarh: have you guys looked in to lustre? |
| 08:46.34 | hali | exporting 32Tb in a TPC env as NFS seems really sub-optimal |
| 08:47.12 | hali | i would export one or two 4Tb luns to each lustre data server |
| 08:47.39 | hali | plus the devs can do app native io calls without bothering the kernel with lustre |
| 08:50.36 | *** join/#gllug stu_ (n=stu@dyn1240-64.vpn.ic.ac.uk) |
| 08:58.45 | antiphase | What's suboptimal about it? |
| 08:58.59 | antiphase | Using some half-finished software to stitch bits of storage together is worse |
| 08:59.43 | hali | lustre is pretty solid |
| 09:00.44 | antiphase | But an extra layer of something for nothing |
| 09:00.59 | hali | it's the same number of layers |
| 09:01.02 | hali | -nfs +lustre |
| 09:01.03 | antiphase | Distributed storage is a pain in the balls |
| 09:01.28 | hali | i've never seen "good" throughput to NFS |
| 09:01.42 | hali | pNFS possibly, but that is sort of distributed storage as well |
| 09:02.11 | hali | even to netapp filers with 10gE cards i haven't seen more than 2-3Gb out of a single head |
| 09:02.13 | antiphase | What's "good"? |
| 09:02.28 | cpufreak | lustre is a bad diea. |
| 09:02.47 | cpufreak | does it even work with a modern kernel yet? |
| 09:03.01 | hali | lustre is used by probably 75% of all linux installments in the topp500 HPC list |
| 09:03.12 | cpufreak | HPC != production. |
| 09:03.13 | antiphase | Are you doing HPC? |
| 09:03.17 | hali | bilarh is |
| 09:03.27 | cpufreak | GPFS is used by lots of HPC stuff, its not production ready. |
| 09:03.34 | hali | kernel version wise not sure, i 've only used it on normal centos installs |
| 09:03.37 | antiphase | They're only crunching a couple of numbers and don't want to pay for commercial software |
| 09:04.41 | hali | cpufreak: did you ever test bluearc boxes over at joost? |
| 09:04.54 | cpufreak | yes |
| 09:05.00 | hali | verdict? |
| 09:05.00 | cpufreak | I wouldn't buy bluarc |
| 09:05.06 | cpufreak | not because of testing them |
| 09:05.08 | cpufreak | they performed ok there |
| 09:05.19 | hali | the price tag is pretty hairy |
| 09:05.29 | cpufreak | nah, price wise it was the 2nd chespest option |
| 09:05.41 | cpufreak | its more the fact as lots of stuff is in hardware |
| 09:05.50 | cpufreak | the normal cpu is pants |
| 09:06.01 | cpufreak | so when something happenes where it falls out of hardware into software |
| 09:06.13 | cpufreak | performance falls off the end |
| 09:06.17 | cpufreak | er edge |
| 09:07.19 | antiphase | AIUI Claranet have had some interesting experiences with storage after moving from NetApp (flawless) to Bluearc (don't ask) |
| 09:07.45 | antiphase | Bluearc stuff basically isn't finished, but they're already selling them |
| 09:08.01 | antiphase | Hence it's cheap, because it might break at any moment |
| 09:08.16 | cpufreak | I don't really see a valid argument for not buying netapp |
| 09:08.39 | hali | yeah, i tend to buy netapp for almost everything |
| 09:08.51 | hali | except distributed hpc envs obviously ;-) |
| 09:09.16 | hali | we added deduplication to our vmware vmstore filers and disk usage went down by about 25% in no time |
| 09:18.36 | cpufreak | don't use vmswear |
| 09:18.37 | cpufreak | too slow |
| 09:18.40 | cpufreak | too much latency |
| 09:18.43 | cpufreak | kvm 4tw |
| 09:19.42 | hali | esx is too easy |
| 09:20.11 | hali | we use vmware more for HA than for the benefit of virtualization |
| 09:21.02 | hali | mostly because some of the in house apps i host can't be load balanced and can only run on a single node at a time bla bla |
| 09:30.00 | *** join/#gllug pepo_ (n=pepo@cpc1-acto10-0-0-cust96.brnt.cable.ntl.com) |
| 09:31.13 | z00dax | cpufreak: at the moment thought, Xen > KVM as far as I can tell - and on my test machine, amd64 6000+ |
| 09:31.44 | z00dax | both in terms if in VM performance ( as measured using tsung from a diff box ) and density |
| 09:31.52 | z00dax | KVM wins on the this-is-good factor though |
| 09:33.43 | hali | i have two citrix xenserver boxes and performance on those is pretty good, especially disk io |
| 09:33.55 | hali | they are hitting a basic nas via iscsi |
| 09:44.07 | cpufreak | z00dax: I think it depends how much network stuff there is |
| 09:44.56 | z00dax | true |
| 09:45.02 | z00dax | in my case, this was benchmarking a webserver |
| 09:47.30 | z00dax | hali: on a dual dual core ( so 4 ), hitting a 12 disk raid10, i can get just over 2.8 million page deliveries/hr out of 8 drupal installs, each in its own Xen VM/512MB ram |
| 09:47.41 | z00dax | which is, iirc, just under 100hits/sec/vm |
| 09:48.34 | z00dax | yes, its test data, so a few hundred webpages in each instance, and running their own mysql's |
| 09:48.58 | z00dax | waiting for a suiteable 'outage', so i can trial this under kvm as well. |
| 09:49.09 | z00dax | lets see what el5.4 has, otherwise will just goto kvm-upstream |
| 10:47.47 | *** join/#gllug celesteh (n=celesteh@sblug/member/celesteh) |
| 11:08.42 | *** join/#gllug Leeds (n=richardc@n058152141045.netvigator.com) |
| 11:36.19 | *** join/#gllug morsing (i=morsing@emil.morsing.cc) |
| 11:36.23 | morsing | antiphase: Where's bilarh? |
| 11:36.32 | antiphase | Round your Mum's |
| 11:36.40 | antiphase | He's a bit of a goer |
| 12:08.31 | *** join/#gllug dick_turpin (n=dick_tur@host217-34-163-30.in-addr.btopenworld.com) |
| 12:08.35 | morsing | dick_turpin! |
| 12:08.40 | morsing | dick_turpin: Where's bilarh? |
| 12:08.54 | dick_turpin | In my cellar |
| 12:11.12 | morsing | Cawfy |
| 13:23.50 | z00dax | check it! |
| 13:26.09 | morsing | Beer |
| 13:27.04 | z00dax | speaking of which, is there a gllug outing planned for beer fest ? |
| 13:31.22 | *** join/#gllug stu_ (n=stu@dyn1240-64.vpn.ic.ac.uk) |
| 13:32.14 | stu_ | anyone have any ideas where i can get a load of free linux promotional stuff? I've already asked Canonical for a load of CDs and linuxemporium. |
| 13:32.24 | stu_ | Its for a freshers fair |
| 13:32.42 | stu_ | fayre? |
| 13:36.11 | antiphase | You really need beer if you want to pique people's interest |
| 13:43.28 | stu_ | antiphase, lol, i think the college might have something to say about that. "use linux and get free beer" |
| 13:45.05 | *** part/#gllug dick_turpin (n=dick_tur@host217-34-163-30.in-addr.btopenworld.com) |
| 14:25.35 | morsing | stu_! |
| 14:38.55 | wethrin | Freshers' Fairs tend not to approve of giving away beer |
| 14:42.58 | cpufreak | free beer outside |
| 14:50.03 | yaMatt | they would much rather charge you for the beer |
| 14:50.53 | morsing | Beer |
| 14:51.05 | morsing | yaMatt: Where's Tarragon? |
| 14:51.26 | yaMatt | who? |
| 14:53.34 | morsing | ~seen Tarragon |
| 14:53.35 | ibot | tarragon <n=tutti@cpc5-stkp5-0-0-cust82.manc.cable.ntl.com> was last seen on IRC in channel #gllug, 211d 17h 25m 34s ago, saying: 'Awful. On call again. Can't believe how much I hate it.'. |
| 14:55.01 | yaMatt | that was a long time ago |
| 16:16.58 | *** join/#gllug cityLights1 (n=cityLigh@bzq-84-108-42-233.cablep.bezeqint.net) |
| 16:17.07 | cityLights1 | anyone here knows alsa? |
| 16:17.25 | cityLights1 | how can I prevent a usb sound card from hijicking card0 in alsa? |
| 16:17.40 | cityLights1 | I tried to fix it with the index at /etc/asound.conf - but it doenst work |
| 17:55.37 | *** join/#gllug jason_ukfsn (n=jason@84.45.232.225) |
| 17:55.46 | jason_ukfsn | hi all |
| 17:55.55 | wethrin | Hey jason |
| 17:55.56 | wethrin | LTNS |
| 17:56.17 | jason_ukfsn | yeah, I've been very busy and had a few health issues too |
| 17:56.25 | jason_ukfsn | how's life? |
| 17:56.27 | wethrin | Mmm. Are you better now? |
| 17:56.33 | wethrin | Life's okay. Keeping as busy as I can do |
| 17:57.10 | jason_ukfsn | feeling better but that's probably because I've finally delivered everything to my accountant to enable her to complete and file the UKFSN accounts |
| 17:57.21 | wethrin | Ah. Glad to hear :) |
| 17:57.35 | wethrin | I'm still stuck in the lab. |
| 17:57.41 | wethrin | So I should do something about this |
| 17:58.13 | jason_ukfsn | probably. I've been working from home the past couple of days |
| 17:58.30 | jason_ukfsn | my son is sitting behind me playing mario smash on the wii |
| 17:58.36 | jason_ukfsn | bloody annoying that is! |
| 17:58.42 | wethrin | Ahh. I got myself a Wii a couple of weeks ago |
| 17:58.43 | wethrin | Great fun! |
| 17:59.24 | jason_ukfsn | It is a lot of fun. The first true revolution in gaming in many years. |
| 17:59.36 | wethrin | Well, possibly. |
| 17:59.39 | wethrin | The controller, sure. |
| 18:00.08 | jason_ukfsn | UKFSN has 2 wii's in their boxes ready for me to launch a new referral rewards system |
| 18:00.09 | wethrin | But the games - I find the games are much simpler than on the, say, Xbox or PS3. |
| 18:00.36 | wethrin | Which, I think, is one of the reasons the Wii is so popular |
| 18:00.47 | wethrin | It's not a system for 'gamerz' |
| 18:00.50 | jason_ukfsn | not just the controller - it's the whole paradigm of using the whole body to control the game - promised by so many VR fantasies but delivered rather elegantly in the wii I think |
| 18:01.00 | wethrin | Got the balance board for it? |
| 18:01.21 | jason_ukfsn | yes. I bought the wii fit thing with the intention of using it regularly which I did for about 2 weeks. |
| 18:01.31 | jason_ukfsn | My wife and kids are still using it so it's not a waste. |
| 18:01.54 | wethrin | I've just been a bit too tired recently to use it |
| 18:02.21 | wethrin | Aanyway. I really must get home and get some food |
| 18:02.26 | jason_ukfsn | I think the focus on producing games to take advantage of the human interface should pay off in getting lots of people gaming who would never have otherwise. |
| 18:02.34 | jason_ukfsn | OK, enjoy |
| 18:02.45 | wethrin | Good to see you're still around, anyhow! |
| 18:02.57 | wethrin | catch you later! |
| 18:34.19 | *** join/#gllug shai (n=Shai@l192-117-110-233.cable.actcom.net.il) |
| 18:43.35 | wethrin | rarr |
| 18:46.36 | morsing | shai! |
| 18:54.39 | jason_ukfsn | boo |
| 19:02.51 | morsing | jason_ukfsn!!!!!!!!!!!!!!!!!! |
| 19:15.01 | jason_ukfsn | yes, that would be me |
| 19:27.16 | wethrin | Well, the curry was tasty. |
| 19:27.35 | morsing | ... |
| 19:41.34 | jason_ukfsn | OK, I've proved I still exist. |
| 19:41.59 | jason_ukfsn | I was going to joke that it's time for me to vanish but I see the owner of a certain ISP has beaten me to that one ;) |
| 19:47.39 | wethrin | Hehe. At least you're not also running a rather major project :) Are you? |
| 19:48.27 | jason_ukfsn | neither is he ;) |
| 19:48.36 | wethrin | well, not any more :) |
| 19:48.55 | jason_ukfsn | The only major project I am running is my daughters transition to secondary school |
| 19:50.17 | jason_ukfsn | I have to say I'm quite shocked at what seems to be happening with CentOS. Not overly surprised but shocked nonetheless |
| 19:51.40 | wethrin | Well, arguably the CentOS project should've got around to setting up a project Paypal and Google Ads account, not relying on Lance |
| 19:51.56 | wethrin | so there seems to be a bit of poor management on that front |
| 19:52.07 | *** join/#gllug DiscordianUK (n=ch@78.146.172.73) |
| 19:52.21 | *** part/#gllug morsing (i=morsing@emil.morsing.cc) |
| 19:53.07 | jason_ukfsn | To do that they would have had to establish some kind of official existence. Lance had kind of done that but it seems that he may be the whole of it and he has a company registered to take ownership in his own name |
| 19:55.47 | jason_ukfsn | oh well, I'm sure they will find some solution to the problem. Rebranding isn't easy but it is doable. |
| 19:56.02 | wethrin | Do they actually need to rebrand? |
| 19:56.07 | wethrin | Well, other than the domain |
| 19:56.45 | jason_ukfsn | the domain is a fairly significant thing for a free software project |
| 19:57.26 | wethrin | true |
| 19:57.27 | DiscordianUK | Good evening |
| 19:57.39 | wethrin | but, say, centos-project.org rather than centos.org would work |
| 19:57.45 | jason_ukfsn | A certain person has a history of reacting badly when his actions (or lack of them) are exposed so I'd hope those actually running the project don't risk it |
| 19:58.58 | jason_ukfsn | So long as centos.org doesn't get turned to something else that might be OK but the only way they could safely continue to use the centos name would be to establish that they have clear ownership of it. There is a risk Lance might claim it belongs exclusively to him as he has control over pretty much everything already |
| 19:59.32 | wethrin | Mmm. But that's not good for a project. |
| 20:00.03 | jason_ukfsn | oh well. It's someone else's problem and my wife wants this desktop so I'll see ya around |
| 21:11.13 | *** join/#gllug Leeds (n=richardc@n058152141045.netvigator.com) |
| 23:29.56 | *** join/#gllug celesteh (n=celesteh@sblug/member/celesteh) |