Let me introduce myself
by Serghey Rodin
Hello guys,
my name is Serghey Rodin and I am from capital of Moldova, Chishinau. I
would like to be a part of fedora-project heart. Im not sure what to say
here :) I like Fedora and I like GNU/Linux very much:)
Regarding my working experience: I worked in best Moldovian Internet
Provider (StarNet http://starnet.md), in outsourcing company Remsys
(http://remsys.net), Web Development company Ceonex (http://ceonex.com).
I have big experience in building failover clusters, security stuff,
monitoring huge networks and so.
Here is some of my skills:
[ OS types ]
Fedora, RHEL, CentOS, Debian, SuSE
FreeBSD
[ Web Servers]
Nginx
Apache
Lighttpd
LSWS
[ DB ]
PostgreSQL
MySQL
[ Languages ]
C (not so good, but I can write something small if its needed)
PHP / HTML
CGI / PERL (not so good, actually I can only troubleshoot code)
BASH (Im a guru :)) )
SQL
[ Hosting Panel's and VPS software ]
OpenVZ
Virtuozzo
Xen
Vserver
HSP/Plesk
cPanel
DirectAdmin
[ Network Equipment ]
Cisco
Foundry
3Com
Dell
Alteon
Thats not all, I'm just trying to show that I can be helpfull.
I will automate just about everything: backups, OS install, system
updates, etc... I am natural multi-tasker I can handle interrupts while
fluidly switching between several projects
Please feel free to ask any questions.
Thanks
--
With Best Regards
Serghey Rodin
16 years, 5 months
Php why must your apps suck so?
by Mike McGrath
Ok, so there's a ticket for a new news site,
https://hosted.fedoraproject.org/projects/fedora-infrastructure/ticket/178
Fact: PHP apps have a poor track record.
Fact: There doesn't appear to be any viable Python CMS's (you can say
plone, but we've managed to fail at installing it for 2 YEARS now. And
at this point I'm not interested in deploying it for anything other than
the docs site)
So lets get this thread started, Toshio ran some numbers to back up this
claim, 73 vulnerabilities since 2006. So what do we do. I'm pretty in
the middle of this debate at this point. Especially with something like
mod_security.
Discuss, comment, flame. Lets go :)
-Mike
16 years, 5 months
Another Xen guest lockup
by Jesse Keating
At some point xen4 went down last night. 22:37 koji time was the last
time the guest running on it (xenbuilder2) checked in to koji. It was
building sbcl, which I think was the thing that was building last time
a builder went down. Perhaps this package is triggering something bad?
--
Jesse Keating
Fedora -- All my bits are free, are yours?
16 years, 5 months
Meeting Log - 2007-11-01
by Ricky Zhou
15:55 * jima runs and hides
15:58 -!- fchiulli [i=824c4010(a)gateway/web/cgi-irc/ircatwork.com/x-a6b81cd860eb6651] has joined #fedora-meeting
16:00 < jima> AIEEE NOOO
16:00 -!- mmcgrath changed the topic of #fedora-meeting to: Infrastructure -- Who's here?
16:00 < mmcgrath> Ping
16:00 < mmcgrath> who's around?
16:01 < markhsa> Mark H
16:01 < paulobanon_> me
16:01 < markhsa> me
16:01 < f13> I'm here
16:01 < fchiulli> fchiulli is listening
16:01 * jeremy is around-ish
16:01 * quaid is lurking
16:01 < bklemm> bklemm is listening in too
16:01 * jima imagines fchiulli and bklemm are new to irc :)
16:02 * jima
16:02 < mmcgrath> spevack: abadger1999 poelcat mbacovsk ricky ivazquez warren skvidal lmacken dgilmore: ping
16:02 < fchiulli> Give that man a cigar.
16:02 < mmcgrath> I know some of them won't be in today.
16:02 < skvidal> pong
16:02 < abadger1999> pong
16:02 < warren> pong
16:02 < poelcat> mmcgrath: pong
16:02 < jima> dgilmore was busy as heck earlier, not sure about now.
16:03 -!- giarc_w [i=hidden-u(a)gnat.asiscan.com] has joined #fedora-meeting
16:03 < lmacken> pong
16:03 < mmcgrath> Welp, lets get started then.
16:03 -!- mmcgrath changed the topic of #fedora-meeting to: Infrastructure -- Release is go!
16:03 < mmcgrath> .tiny https://hosted.fedoraproject.org/projects/fedora-infrastructure/query?sta...
16:03 < mmcgrathbot> mmcgrath: http://tinyurl.com/292qq8
16:04 < mmcgrath> So yeah, AFAIK we will release Fedora 8 on the 8th next week.
16:04 < mmcgrath> Is anyone confused about what they should or shouldn't be doing at this point?
16:04 < jima> should: looking for bugs
16:04 < spoleeba> mmcgrath, duck and cover?
16:04 < warren> meh
16:04 < jima> shouldn't: changing infrastructure without asking :)
16:05 < skvidal> jima: wait? what?
16:05 < skvidal> should NOT
16:05 < jima> spoleeba: i thought we did that every day ;)
16:05 < skvidal> oh shit
16:05 < abadger1999> :-)
16:05 * mmcgrath smacks skvidal
16:05 -!- nim-nim [n=nim-nim@fedora/nim-nim] has quit "Leaving."
16:05 < skvidal> mmcgrath: hey, I wasn't the one who brought proxy4 online yesterday! :)
16:05 < mmcgrath> Just a reminder, ANOTHER one :) After tonight, no changes to infrastructure without contacting the list first.
16:05 < jima> skvidal: ...?
16:05 < skvidal> jima: I'm making fun of mmcgrath
16:06 < mmcgrath> jesse also had a wish to extend this period for next release. I'm generally in favor of that though I'd like to see hwo this change freeze goes.
16:06 * dgilmore is sitting on the cold floor of the colo
16:06 < mmcgrath> There are a few critical things I'd like to talk about.
16:06 -!- mmcgrath changed the topic of #fedora-meeting to: Infrastructure - mmcgrath (schedule)
16:06 < jima> hmm, should i be getting approval to work on ticket #220? :)
16:06 < mmcgrath> jima: lets get to that in a bit
16:06 < jima> 'k
16:06 -!- nim-nim [n=nim-nim@fedora/nim-nim] has joined #fedora-meeting
16:07 < mmcgrath> So, if I've done my job right, I won't need to be around for this release. Having said that. I'm going to try to be.
16:07 < paulobanon_> no changes no fun!
16:07 < mmcgrath> skvidal is going to be the point person during the release, even if I am around.
16:07 < skvidal> mmcgrath: do not be around
16:07 < skvidal> seriously - don't do it
16:07 < mmcgrath> :)
16:07 < jima> skvidal: see, you shouldn't have made fun of him -- he just got payback.
16:07 < skvidal> we'll deal w/o you :)
16:07 < mmcgrath> So here's my schedule.
16:07 < skvidal> jima: I [was] volunteered
16:08 < mmcgrath> I'm off tomorrow and will largely be unavailable until Monday.
16:08 < jima> skvidal: exactly my point. ;)
16:08 < jima> mmcgrath: have fun! :)
16:08 < mmcgrath> I'll be taking half days Monday - Thursday (release day) and on the 9th I'll be gone.
16:08 < mmcgrath> just gone. Like I wasn't ever here.
16:08 < jima> damn ninja.
16:08 < mmcgrath> and I'll be back for good on the 13th.
16:08 < jeremy> who's mmcgrath?
16:08 < skvidal> mmcgrath: wow, like keyser sose
16:08 < mmcgrath> Hopefully with a tan. Though its likely I'll burst into flames the first time I come in contact with the sun.
16:09 < skvidal> KEYSER SOSE KEYSER SOSE!
16:09 < mmcgrath> :)
16:09 < jima> i think anything bursts into flame when it comes even *close* to the sun.
16:09 < mmcgrath> So mark it down and remember that we talked about it in the meeting.
16:09 < mmcgrath> Next topic!
16:09 < paulobanon_> wedding topic ?!
16:09 -!- mmcgrath changed the topic of #fedora-meeting to: Infrastructure -- Current nown oddities.
16:09 -!- mmcgrath changed the topic of #fedora-meeting to: Infrastructure -- Current known oddities.
16:09 < mmcgrath> damn
16:09 < mmcgrath> Ok.
16:09 < jima> known oddity: "nown" is missing a 'k'
16:09 < skvidal> so, next week after we remove all of mmcgrath's accounts what will we be doing
16:09 < skvidal> ?
16:09 < skvidal> oh
16:09 < skvidal> damn
16:10 < skvidal> wrong channel
16:10 < jima> skvidal: that was for #fedora-evil, huh?
16:10 < mmcgrath> heh
16:10 < skvidal> it was
16:10 < mmcgrath> https://hosted.fedoraproject.org/projects/fedora-infrastructure/ticket/222
16:10 < mmcgrath> .ticket 222
16:10 < mmcgrathbot> mmcgrath: #222 (sysctl on the proxy servers) - Fedora Infrastructure - Trac
16:10 < mmcgrath> I should have had this done a week ago but testing it and being busy with other stuff put it off.
16:10 < mmcgrath> Long story short.
16:10 -!- stahnma_ [n=stahnma(a)c-76-18-178-254.hsd1.tn.comcast.net] has joined #fedora-meeting
16:10 < yingbull> I should probably say hi - I was around earlier in the fall, but got a bad case of the $DAYJOB, and had fedora account issues that slowed me down getting involved (and without anything tasked to me yet, I just sorta faded off). Back around now, tho. So... hi. ;-)
16:10 < mmcgrath> The sysctl configuration optimizations have been done on all the proxy servers.
16:10 < mmcgrath> yingbull: welcome!
16:11 < yingbull> made it to this meeting at least, hopefully its a good start. :-) Sorry to interrupt.
16:11 < mmcgrath> So here's the scoop, we've got optimized proxy servers. I've done all the testing I can to ensure that we are getting peak performance from them.
16:11 * jima waves to yingbull
16:11 < jima> mmcgrath: seems like a clever idea, yeah.
16:11 < jima> is that all the proxies or just 1-2?
16:11 < mmcgrath> Having said that, I don't feel they're tested enough to just deploy. So to mitigate this risk my *optimized* config is in /root/sysctl.conf and the normal config thats been deployed for many months is in /etc/sysctl.conf
16:11 < mmcgrath> jima: all of them.
16:12 < mmcgrath> So if anyone sees the proxy servers getting overwhelmed, and feel it might be tuned wrong, just run "sysctl -p"
16:12 < mmcgrath> if you want to go back to the 'optimized configs' run "sysctl -p /root/sysctl.conf"
16:12 < mmcgrath> profit.
16:12 * jima almosts asks "where's a copy of the optimized one in puppet?" then remembers mmcgrath provided us with a copy.
16:12 < mmcgrath> Thats an odditiy I thought I'd mention.
16:12 -!- clarkbw [i=clarkbw@nat/redhat/x-ae3597e3eeab15df] has joined #fedora-meeting
16:12 < mmcgrath> since it doesn't make any obvious sense.
16:13 -!- londo [n=georgiou(a)82.133.49.59] has quit Read error: 104 (Connection reset by peer)
16:13 -!- clarkbw [i=clarkbw@nat/redhat/x-ae3597e3eeab15df] has quit Read error: 104 (Connection reset by peer)
16:13 < mmcgrath> Anyone have any questions on that? Did anyone see anything obviously wrong with the sysctl.conf I sent to the list?
16:13 < lmacken> looked sane to me
16:13 < jeremy> seem fine to me
16:13 < mmcgrath> ok, moving on.
16:14 < mmcgrath> Oddity number 2.
16:14 -!- rdieter is now known as rdieter_away
16:14 < mmcgrath> The wiki notifications today.
16:14 < mmcgrath> We've been seeing higher than normal wiki notifications as of today.
16:14 < mmcgrath> I don't think the wiki is any worse off today then it was yesterday. I think its because we've been monitoring the wiki with caching enabled and I recently made that cache more aggressive than normal.
16:15 < mmcgrath> The notifications we're getting are from the only non-cached page.
16:15 < mmcgrath> Which means A) the wiki is unresponsive at times.
16:15 < mmcgrath> but B) I've not witnessed this myself, only seen the nagios alerts.
16:16 < mmcgrath> I'm watching this closely and trying to tune the wiki a bit but so far nothing has made the load any better.
16:16 < mmcgrath> Long story short, the caching I added sees between a 50% and 75% hit rate.
16:16 < jima> that's good.
16:16 < mmcgrath> that means only 50% - 25% of the requests actually hit the app server.
16:16 < mmcgrath> this is something we did not have for the F7 release.
16:16 -!- clarkbw [i=clarkbw@nat/redhat/x-29205887cfe9d537] has joined #fedora-meeting
16:16 < mmcgrath> The bulk of our work will be done by the proxy servers this time around.
16:17 < jima> but god help the actual wiki for the hits that do get through. :|
16:17 < mmcgrath> of which we have 4, the 2 new ones have more ram and are x86_64 boxes. So I think we're in good shape.
16:17 < mmcgrath> The wiki, as always, concerns me.
16:17 -!- k0k [n=k0k@fedora/k0k] has quit Read error: 110 (Connection timed out)
16:17 < mmcgrath> It just does.
16:17 * jima concurs
16:17 < mmcgrath> I've moved app3 off of xen1 and on to xen2 to dedicate more resources to xen1 (which currently contains the wiki and one proxy server)
16:17 < jeremy> we should probably just be aggressive about making "important" pages static and for the rest, if the wiki goes down, the wiki goes down
16:17 < poelcat> skvidal: is the problems I had yesterday app1 and with updating wiki pages is fixed?
16:17 < mmcgrath> unfortunately thats the best setup I've found so far.
16:18 < skvidal> poelcat: for certain values of fixed, yes
16:18 < mmcgrath> Any questions about the wiki right now?
16:19 < jima> mmcgrath: yes, but my questions keep timing out.
16:19 < mmcgrath> I've also turned off the translate app for app1 for now.
16:19 * mmcgrath smacks jima.
16:19 < mmcgrath> especially over the next many days. If you have ANY problems with the wiki, please email me the time, what you were trying to do and your IP address.
16:20 < mmcgrath> Ok, moving on.
16:20 < mmcgrath> mdomsch is not around but all that is left on that end is to disable the primary mirrors initially (download.fedora.redhat.com) from the mirror list and public list.
16:20 * jima scratches head
16:21 -!- mmcgrath changed the topic of #fedora-meeting to: Infrastructure -- Specific tickets open.
16:21 < jima> disable them from mirrorlist? isn't that tweakable via mirrormanager?
16:21 < mmcgrath> .ticket 93
16:21 < mmcgrathbot> mmcgrath: #93 (Fedora 8) - Fedora Infrastructure - Trac
16:21 < mmcgrath> We'll close that when everything is done
16:21 < mmcgrath> .ticket 97
16:21 < mmcgrathbot> mmcgrath: #97 (Fedora 8 Website) - Fedora Infrastructure - Trac
16:21 < mmcgrath> Ricky and I have been working on this together, expect to see the F8 website by the end of the day but with the F7 graphics and such.
16:21 < mmcgrath> We'll have a special build for F8 for the mirrors by the end of the day on Monday.
16:21 < mmcgrath> .ticket 99
16:21 < mmcgrathbot> mmcgrath: #99 (Mirror Coordination) - Fedora Infrastructure - Trac
16:22 < mmcgrath> Currently assigned to mdomsch either he or jesse keating will likely notify the mirrors when the bits are ready.
16:22 < mmcgrath> .ticket 100
16:22 < mmcgrathbot> mmcgrath: #100 (Web Content on the mirrors) - Fedora Infrastructure - Trac
16:22 < mmcgrath> already discussed, this will be done with ticket #97
16:22 < mmcgrath> .ticket 101
16:22 < mmcgrathbot> mmcgrath: #101 (Fedora Infrastructure Change Freeze) - Fedora Infrastructure - Trac
16:22 < mmcgrath> I'll close this ticket AFTER tonight.
16:22 < mmcgrath> ticket .137
16:22 < mmcgrath> .ticket 137
16:22 < mmcgrath> :)
16:22 < mmcgrathbot> mmcgrath: #137 (Sponsorship prominence) - Fedora Infrastructure - Trac
16:23 < mmcgrath> This is actually in good shape and live. Its linked at the bottom of the page http://fedoraproject.org/sponsors
16:23 < mmcgrath> .ticket 210
16:23 < mmcgrathbot> mmcgrath: #210 (Disable F8 in Mirrormanager) - Fedora Infrastructure - Trac
16:23 < notting> g
16:24 < mmcgrath> This is also currently assigned to mdomsch (who's not here) but I'll follow up with him to make sure this is done correctly when the bits go
16:24 < mmcgrath> .ticket 211
16:24 < mmcgrathbot> mmcgrath: #211 (Enable Bodhi Updates) - Fedora Infrastructure - Trac
16:24 < mmcgrath> lmacken: ?
16:24 < mmcgrath> This is assigned to you, still all good?
16:24 < lmacken> shouldn't be a problem
16:24 < mmcgrath> any ETA?
16:25 < lmacken> it's just a bitflip.. whenever we want to unleash
16:25 -!- stahnma_ is now known as stahnma
16:25 < mmcgrath> k, I'll leave that up to you and releng then :)
16:25 < mmcgrath> .ticket 54
16:25 < mmcgrathbot> mmcgrath: #54 (Postfix Server) - Fedora Infrastructure - Trac
16:25 < mmcgrath> This, unfortunately, will probably be the only ticket we didn't completely finish. Mostly an -ENOTIME and a -EALREADYHAVESOMETHING
16:26 < mmcgrath> bastion's been fine, this was a bigger priority when the mailman list stuff was being discussed and we had OTRS.
16:26 < mmcgrath> and the last ticket
16:26 < mmcgrath> .ticket 136
16:26 < mmcgrathbot> mmcgrath: #136 (Translated Website) - Fedora Infrastructure - Trac
16:26 < mmcgrath> We've 'frozen' the website and the translation team is currently at work on it
16:26 < jima> yay for translators!
16:26 * mmcgrath notes
16:26 < mmcgrath> http://translate.fedoraproject.org/module/fedora-web
16:27 < mmcgrath> I think this is actually a big deal.
16:27 < mmcgrath> but the final 'freeze' will be on monday night.
16:27 < mmcgrath> So really, we're in good shape.
16:27 < mmcgrath> Hopefully we won't have to juggle much at release time.
16:27 < mmcgrath> Anyone have any questions or concerns regarding me, our infrastructure, stuff not staying up or getting overloaded?
16:28 < mmcgrath> skvidal: you've got to have questions about stuff right?
16:29 < skvidal> mostly i'm reading
16:29 -!- EvilBob [n=bob(a)fedora/pdpc.sustaining.BobJensen] has quit Remote closed the connection
16:29 * mmcgrath also notes http://fedoraproject.org/wiki/Infrastructure/SOP/Release
16:29 < mmcgrath> still exists and is valid.
16:29 < skvidal> but right now my primary thoughts are making sure we have all the steps down for dynamically making static pages when we need them
16:29 < mmcgrath> skvidal: we should make an SOP for that.
16:30 < mmcgrath> skvidal: OH, thats actually a good point.
16:30 < skvidal> but we can't get to the SOP :)
16:30 < skvidal> so it's better just to have a note
16:30 < skvidal> so i'm thinking
16:30 < mmcgrath> *if* the wiki gets overloaded and you can't get to it, log into RH VPN and go to app2.fedora.phx.redhat.com/wiki/
16:30 < mmcgrath> since we can't balance the wiki its always on one or the other, just go to the other :)
16:30 < skvidal> 1. save the page in firefox
16:30 < mmcgrath> skvidal: do it with wget.
16:30 < skvidal> 2. upload the files to the static path on the proxies
16:31 < skvidal> wget, in the past, didn't grab css for me
16:31 < mmcgrath> hmm, k, we can test that after the meeting.
16:31 < skvidal> okie doke
16:31 < mmcgrath> I'd also encourage everyone to become familiar with our caching setup. especially the headers.
16:32 < mmcgrath> .headers http://fedoraproject.org/wiki/
16:32 < mmcgrathbot> mmcgrath: apptime: D=77335, proxytime: D=167438, expires: Thu, 01 Nov 2007 21:32:01 GMT, vary: Cookie,User-Agent,Accept-Language, connection: close, server: Apache/2.2.3 (Red Hat), appserver: (null), proxyserver: proxy3.fedoraproject.org, date: Thu, 01 Nov 2007 20:32:01 GMT, content-type: text/html; charset=utf-8
16:32 * mmcgrath notes proxy and app time
16:32 < mmcgrath> .headers http://fedoraproject.org/wiki/
16:32 < mmcgrathbot> mmcgrath: apptime: D=77335, content-length: 13644, proxytime: D=87, expires: Thu, 01 Nov 2007 21:32:01 GMT, vary: Cookie,User-Agent,Accept-Language, connection: close, server: Apache/2.2.3 (Red Hat), appserver: (null), proxyserver: (null), date: Thu, 01 Nov 2007 20:32:15 GMT, content-type: text/html; charset=utf-8, age: 15
16:32 * mmcgrath notes "age" this time.
16:32 < mmcgrath> stuff like that is important for understanding why USER A) gets cached copy and why user B) doesn't.
16:33 < mmcgrath> If you want to flush all the cache, just "rm -rf /srv/web/cache/*" on the proxy box.
16:33 -!- londo [n=georgiou(a)82.133.49.59] has joined #fedora-meeting
16:33 < mmcgrath> blah blah.
16:33 < mmcgrath> Ok, if no one has any questions about anything else I'll open the floor?
16:33 < skvidal> mmcgrath: one more question
16:33 < skvidal> on the proxy boxes
16:33 < skvidal> we have /srv/web/static-web
16:34 < skvidal> is that path old/broken?
16:34 < mmcgrath> That is the old path.
16:34 * mmcgrath notes proxy4
16:34 < mmcgrath> it doesn't have it.
16:34 < skvidal> right - I was more thinking of cleaning that up
16:34 < mmcgrath> <nod>
16:34 < mmcgrath> I thought I'd moved that to /tmp/
16:35 < mmcgrath> perhaps I moved it back during the conversion.
16:35 < mmcgrath> but yeah, that should now be "/srv/web/fedoraproject.org/" or "/srv/web/start.fedoraproject.org/"
16:35 < skvidal> looks like you copied it to temp :)
16:35 < skvidal> but it never got nuked
16:35 * mmcgrath also notes that AFAIK, search.fedoraproejct.org won't be ready.
16:36 < mmcgrath> but thats not really our concern, just pointing it out.
16:36 < yingbull> my bit: I'll want something to do soon. I need to fix my fedora account, but will want something to work on to get started. Else I'm likely to let $DAYJOB eat my $NIGHTLIFE, and without anything depending/waiting on me, fade away again. :-) Anyone got something that I can thing them about once my fedora account is working?
16:36 < mmcgrath> yingbull: how's your mysql skills?
16:36 -!- clarkbw [i=clarkbw@nat/redhat/x-29205887cfe9d537] has quit Read error: 110 (Connection timed out)
16:37 < yingbull> mmcgrath: not bad, haven't had a huge need for it lately, but I did work with it before.
16:37 < yingbull> Used it for a big web-app at the last place I worked.
16:37 < mmcgrath> yingbull: ping me after the meeting in #fedora-admin, we'll find something definitive for you.
16:37 < yingbull> Will do. Save me from $DAYJOB, please! :-)
16:37 < mmcgrath> heh
16:37 < mmcgrath> ok
16:37 -!- mmcgrath changed the topic of #fedora-meeting to: Infrastructure -- Open Floor
16:38 < mmcgrath> Does anyone have anything specific they'd like to discuss?
16:38 < mmcgrath> I'd encourage everyone to start thinking of F9 goals.
16:38 * jima muses
16:39 < mmcgrath> Anyone have anything else? If not we'll close the meeting.
16:39 < jima> mmcgrath: ticket 220?
16:40 < mmcgrath> .ticket 220
16:40 < mmcgrathbot> mmcgrath: #220 (nagios monitoring for puppet1) - Fedora Infrastructure - Trac
16:40 < mmcgrath> jima: ahh, yeah I just wanted to make sure it was beign monitored.
16:40 < jima> it's not. next question? :)
16:41 < mmcgrath> well, can we make it monitored? ICMP, disk, puppetmaster, http.
16:41 < mmcgrath> standard stuff.
16:41 < jima> is it running nrpe?
16:41 < mmcgrath> it should be.
16:41 < jima> yep, it is.
16:42 < jima> should i make those changes to nagios? no conflicts with the infrastructure freeze?
16:42 < mmcgrath> jima: I'd put that as a low priority but if you can't get it done tonight, I'd wait until after the release.
16:42 < jima> 'k.
16:42 < nim-nim> ticket 215 anyone?
16:42 < mmcgrath> .ticket 215
16:42 < mmcgrathbot> mmcgrath: #215 (New mailing list request: fedora-fonts-bugs-list) - Fedora Infrastructure - Trac
16:42 < mmcgrath> nim-nim: that one's on me. I'll put that request in now.
16:43 < nim-nim> mmcgrath: thanks!
16:43 * mmcgrath has been a little busy :)
16:43 * nim-nim understands
16:43 < mmcgrath> I'd also like to get the rest of the hosted repos created that are in the queue by tonight.
16:43 < nim-nim> still, nim-nim needs to get his SIG in shape too
16:44 < nim-nim> many thanks!
16:44 -!- JSchmitt [n=s4504kr@fedora/JSchmitt] has quit "Konversation terminated!"
16:44 < mmcgrath> Ok, anyone have anything else?
16:45 < mmcgrath> If not I'll close the meeting in 30.
16:45 < mmcgrath> 15
16:45 < mmcgrath> 5
16:45 < mmcgrath> Good luck to us all next week :)
16:46 -!- mmcgrath changed the topic of #fedora-meeting to: Infrasturcture -- Meeting Closed.
16:46 -!- fchiulli [i=824c4010(a)gateway/web/cgi-irc/ircatwork.com/x-a6b81cd860eb6651] has left #fedora-meeting []
16:46 < mmcgrath> Thanks for coming everyone.
16 years, 5 months
Moin and notifications
by Mike McGrath
I'm just going to throw it out there and see what people say.
Proposal: Disable notifications in Moin
Reason: Even page creations at this point are not working for some people.
-Mike
16 years, 5 months
Sysctl on the proxy servers
by Mike McGrath
I'd like to discuss this at the meeting today, here are the
optimizations as they stand for our proxy boxes. Its ticket #222:
# Kernel sysctl configuration file for Red Hat Linux
#
# For binary values, 0 is disabled, 1 is enabled. See sysctl(8) and
# sysctl.conf(5) for more details.
# Controls IP packet forwarding
net.ipv4.ip_forward = 0
# Controls source route verification
net.ipv4.conf.default.rp_filter = 1
# Controls the System Request debugging functionality of the kernel
kernel.sysrq = 1
# Controls whether core dumps will append the PID to the core filename.
# Useful for debugging multi-threaded applications.
kernel.core_uses_pid = 1
# Ensure connection tracking isn't limiting our connections
net.ipv4.ip_conntrack_max=6553600
# Allow higher than default file descriptors
fs.file-max=4947900
# How many pages to free at a time
vm.page-cluster = 7
# Try to always keep this amount free
vm.min_free_kbytes = 10000
# Allow system to be a swappier than normal when it needs to be for
caching server
vm.swappiness = 60
# Security, protects against TIME WAIT attacks
net.ipv4.tcp_rfc1337 = 1
# Security, protects against SYN floods
net.ipv4.tcp_syncookies = 1
# Lower keep alive time on the edge proxies
net.ipv4.tcp_keepalive_time = 300
# Limit tcp orphans
#net.ipv4.tcp_max_orphans = 1000
# Give the network stack access to more memory for queueing
net.core.rmem_default = 262144
net.core.rmem_max = 262144
16 years, 5 months
Reminder of change freeze
by Mike McGrath
Just a reminder that after tomorrow night no changes are allowed of any
kind in Fedora's infrastructure without discussing it on the list
first. This includes software upgrades, config changes, etc.
Exceptions are outages / issues.
Changes already in the ticketing system related to the release (like
getting the Fedora 8 content generated and put on the website)
-Mike
16 years, 5 months
Introduction to List
by amolarakh
Hello list,
I am Amol Arakh
Basically I am Computer Engineer,and i am end user of RedHat Linux OS
since RedHat 9.0 Launched,.
Currently I am working as Linux System Administrator in
VPCOE,Baramati,Maharashtra,INDIA (An Pronouned Educational Institute)
since last 2-3 years.
In my Network Currently Most of Clients are working on Fedora Core 6.0.
I am proud to say that for the installation purpose I am using one of
Famous Network Installation System named as DRBL,for Network in our
Institute.
I have hands on Practice on Configuration of
IPTABLEs,SQUID,Samba,APACHE,vsFTPD,PAM configuration,Packaging,Taking
Backups.
Thanks,
Warm Regards,
Amol Arakh
16 years, 5 months
