Web   ·   Wiki   ·   Activities   ·   Blog   ·   Lists   ·   Chat   ·   Meeting   ·   Bugs   ·   Git   ·   Translate   ·   Archive   ·   People   ·   Donate

#treehouse, 2010-06-08

Index | Today     Channels | Search | Join

All times shown according to UTC.

Time Nick Message
16:05 dogi #info Weekly Infrastructure meeting:
16:05 #info Volunteer Infrastructure Gang (http://olpcorps.org/ ),
16:05 #info Sugarlabs Infrastructure Team (http://sugarlabs.org/ ),
16:05 #info and TreeHousers (http://me.etin.gs/treehouse/ )
16:05 #info Date: 2010-06-08
16:05 #info Time: 20:00 UTC (16:00 EST, 22:00 CET)
16:05 #info Agenda: http://openetherpad.org/cEp14BO8ov
16:05 #info Location: #treehouse on irc.oftc.net
16:05 #link http://embed.mibbit.com/?serve[…]nnel=%23treehouse
16:05 #topic last meeting
16:05 #info Usefull Links:
16:05 #info LastAgenda:
16:05 #link http://openetherpad.org/wGNTofNahw
16:05 #info LastLog:
16:05 #link http://me.etin.gs/treehouse/tr[…]0100525_1605.html
16:05 #info LastMinutes:
16:05 #link http://me.etin.gs/treehouse/tr[…]0100525_1605.html
16:05 #info NextAgenda:
16:05 #link http://openetherpad.org/md6SQKy996
16:06 #topic agenda
16:06 #info * FSF dns changes and other cleanings
16:06 #info * schooltool and sugar
16:06 #info * jabber.sl.o and further steps
16:06 #info * ex wikipedia servers update
16:06 #info * homework for all treehousers
16:06 #info * access fairy
16:06 are there other topic to disscuss?
16:06 #topic FSF dns changes and other cleanings
16:06 #info fsf changed the dns
16:06 #info sunjammer and all VM on treehouse have to be changed
16:07 silbe can you help me here
16:08 silbe dogi: AFAIK bernie already updated resolv.conf on all VMs
16:08 dogi Can you please update
16:08 /etc/resolv.conf
16:08 to use
16:08 nameserver
16:08 nameserver
16:08 instead of
16:08 nameserver
16:08 thx silbe
16:08 silbe we might need to restart some services if they stopped working because they didn't notice the change
16:09 this probably should mostly affect the Fedora-based VMs as Debian has a fix in glibc and thus Ubuntu should have it as well (but haven't verified that)
16:09 dogi #action dogi look and restart services on treehouseVM
16:09 silbe, +1
16:10 the other change you proposed was to have password less serial login on treehouse servers?
16:10 do you think this is secure enough?
16:11 hi jelkner
16:11 silbe dogi: only root@treehouse can access the (emulated) serial consoles on the VMs
16:11 jelkner hi dogi!
16:11 just read your email reminding me to be here ;-)
16:12 dogi jelkner, backlog: http://me.etin.gs/treehouse/tr[…]0100608_1605.html
16:12 silbe if an attacker made it that far, anything is fair game anyway
16:13 dogi so you think that if an attacker made it into the host all VMs are compromised anyway?
16:13 lfaraone dogi: I'd agree.
16:14 jelkner: ask matt for his opinion.
16:14 jelkner dogi: read the backlog, i don't know what you want me to see?
16:14 dogi would like to have _bernie and hhardy_ s opinion here ...
16:14 lfaraone, thx :)
16:14 jelkner lfaraone, mattva01 is about to join this meeting
16:14 dogi is the same opinion as silbe ...
16:14 lfaraone jelkner: wikimedia servers, I think.
16:16 dogi #idea passwordless login for serial login ...
16:16 silbe hopefully rather #action :-P
16:16 without it I can't look after 90% of the VMs
16:17 (without disrupting the services running on them, that is)
16:17 dogi silbe, :)
16:17 hi mattva01
16:18 mattva01 hey
16:18 dogi mattva01, backlog http://me.etin.gs/treehouse/tr[…]0100608_1605.html
16:18 jelkner dogi, lfaraone, mattva01: we need to figure out what to do with these rack mount servers?
16:18 are they of any use?
16:19 they are taking up space and just sitting here
16:19 dogi jelkner, ok
16:19 #topic ex wikipedia servers updates
16:19 #info RIT has 3 of the 12 exwikipedia servers
16:19 #link http://wiki.sugarlabs.org/go/M[…]iscovery_One/Node
16:19 #info 3 of the 12 are in boston
16:19 #info the last 6 are in Washington DC
16:19 lfaraone jelkner: well, I think the WMF servers you have are all the slower ones.
16:20 jelkner: the three I have are faster, but we're still looking for a home.
16:20 dogi: the RAM for Ivan's server arrived today. I'll be able to bring it into OpSource and into production when I get the HDD etc. Afterwards, i can talk to them about racking the WMF/Devis/whatever servers.
16:21 dogi lfaraone, this are very good news :)
16:22 so let make a count of all the servers which are in virginia ...
16:22 blackrock
16:22 3 wmf servers at lfaraone place
16:22 and an other one in arlington career center
16:23 lfaraone dogi: three at ACC, but two were broken.
16:23 dogi one wmf server was broken ... no working network device
16:23 +1 lfaraone
16:24 so 2 was broken ... I have seen only one
16:25 jelkner, mattva01 was you able to repair more of the dell servers?
16:25 was able to finish one
16:25 lfaraone dogi: the main problem afaict were lack of working fans.
16:26 mattva01 it had more significant issues.....
16:26 dogi lfaraone, how many units is blackrock high?
16:26 lfaraone dogi: two.
16:27 iirc.
16:27 mattva01 even after we got it posting reliably , booting Linux was not reliable due to memory errors
16:28 dogi mattva01, so you had still the same problems as me ...
16:29 mattva01 yep
16:29 dogi think is not a bad quota ... mattva01 we was able to repair one bigger machine for jelkner and one little for sugar :) ... the rest was not so reliable
16:30 so all together there would be 7 units to rack ...
16:31 lfaraone, do you think that is possible ...
16:32 2 units for blackrock
16:32 lfaraone dogi: uh, that sounds like 3 units.
16:32 dogi: unless I'm also racking jeff's machine, which wouldn't make sense for me to do.
16:32 dogi 3 times 1 unit for faster wmf
16:33 1 unit for normal wmf
16:33 1 unit for a  little dell server
16:35 i sorted the machines already by there importance ...
16:35 lfaraone I don't know. I'll have to ask. 2 has been approved. 4 might be reasonable. 7 is pushing it.
16:35 dogi lfaraone, thought jeffs machine will stay in arlington ...
16:35 lfaraone dogi: they will, I was confused.
16:36 dogi :)
16:36 4 machines? or units?
16:36 lfaraone dogi: units.
16:36 dogi: right now they're expecting two.
16:37 dogi hmm
16:38 cause if I could I would rack blackrock and the 3 faster WMF
16:39 lfaraone dogi: I'll see what I can do. As before, we're blocked on the parts for blackrock.
16:39 dogi: I can re-triage and identify the WMF servers we'd want to rack this or next weekend while I'm waiting.
16:39 dogi re-triage?
16:40 lfaraone, which part is missing on blackrock?
16:40 lfaraone dogi: double check whether they all work, identify the better-working ones, run memtest, massage their power supplies, etc. :P
16:40 dogi thx lfaraone
16:41 harddrives?
16:41 lfaraone dogi: that's what we're waiting for, yes.
16:41 dogi ok
16:41 lfaraone dogi: he's got two more 250GB coming iirc.
16:41 dogi :)
16:42 that sounds perfect ...
16:44 was not able to move the boston servers because of holiday weeks around memorial weekend
16:44 my MIT contact returned yesterday
16:45 jlew, thank you for asking cdeslandes for VMs
16:46 #action dogi setup puppet session with dfarning
16:46 jlew, are there updates for RIT?
16:49 jlew nothing that i know of
16:49 dogi there are two other topics open ... but the persons in charge are missing
16:49 hi jlew
16:49 thx
16:49 jlew just got back yesterday, so still figuring things out
16:49 dogi :)
16:50 #topic jabber.sl.o and further steps
16:50 #info jabber.sugarlabs.org is managed by
16:50 #link http://www.solutiongrove.com
16:50 #idea treehouse would like to start a new jabberVM?
16:51 _bernie, yes I am in favor of a VM there ... so give me an ok and I will create one ...
16:53 silbe dogi: why a VM? :-P
16:53 dogi lfaraone, silbe hhardy_ opinions
16:53 :) silbe
16:53 cause jabber is not stable
16:53 or better ejabberd
16:53 lfaraone dogi: sure.
16:54 dogi silbe, http://me.etin.gs/treehouse/tr[…]0525_1605.html#33
16:54 silbe dogi: does it really affect other services?
16:55 dogi silbe, that is a good point
16:55 silbe dogi: I'd only create a VM if a) it needs a special environment or b) is a risk for other services on the same host.
16:56 with b) specifically not including being IO-bound as VMs make that much worse (see SKS keydump import on lightwave).
16:56 _bernie oops
16:56 mmeting
16:56 dogi does not know ... but it is easier to stabilize a service in a VM
16:56 _bernie :-/
16:57 dogi hi _bernie
16:57 _bernie I forgot...
16:57 I'll read the backlog
16:57 silbe dogi: how so?
16:57 _bernie: ok, can stop bernie emulation mode then ;)
16:57 (re. VMs)
17:00 _bernie dogi, silbe: I think root with empty password is fairly safe
17:00 dogi :)
17:00 then we all agree
17:01 silbe _bernie: if PAM is configured correctly I think that's fine
17:01 dogi +1 silbe
17:03 _bernie, by the way are you speaking with dfarning from time to time?
17:05 hmm
17:07 is there somebody else which wants to speak ...
17:08 _bernie, tell me your opinion on jabberVM ...
17:08 #topic homework for every treehouser
17:08 #info "Solve an action per week"
17:08 #info is the main ritual in #treehouse channel
17:08 #action treehouser since an vote on an idea is equivalent to an action ... please vote
17:09 <- this was the last topic :)
17:09 silbe ok, have a nice day everyone :)
17:09 dogi silbe, forgot
17:10 silbe ?
17:10 dogi the experience with pootle and aslo showed me that it was easier to solve the problem in a separate VM
17:12 silbe dogi: why? i.e. what's easier if it's a separate VM?
17:12 the Windows way? :-P
17:13 dogi bingo the stupied old iterative way ...
17:13 :P
17:13 but that is not the only one ... I know this
17:14 #endmeeting

Index | Today     Channels | Search | Join

Powered by ilbot/Modified.