|
evol262 posted:We use Jenkins everywhere, for everything. I'm a bit further along with this now. I have jenkins and gitorious VMs up and running and jenkins is correctly picking up pushes to the git repo. The bit I'm unsure of is how to manage sending changes to the production servers. What I was thinking was: Devs will develop with their local git repo and vms and push changes that are ready for testing to the 'staging' remote which is the one linked to jenkins. This triggers all the unit tests , linters etc, then send the files to our staging webservers for manual testing. What I was unsure about was the best way to get the changes which are ready for deployment onto the production web servers. Most of the DeployOverProtocolX jenkins plugins appear to send every file in the codebase instead of just the changes if I'm understanding it correctly. Is it better to push changes to a 'production' remote git repo instead. e.g. like this? Or is there a better way which is more tightly integrated into jenkins?
|
#
?
Aug 7, 2014 14:45
|
|
|
|
| # ? Apr 29, 2024 18:44 |
|
|
evol262 posted:You don't have a virus. Most Linux exploits are problems with code which become known shortly before after fixes, at which point systems which haven't been updated are vulnerable (Debian/Ubuntu had a problem with breakable encryption keys a few years ago). Viruses exist, but are very rare these days (there were a few in the earlier days of the internet on various UNIX flavors). Thank you for the tips! I didn't see any unrecognized IPs in auth.log, and rkhunter & chrootkit both came up clean, so I guess I'm safe? I deleted my dummy account and so far it doesn't seem to be sending out more spam. Just to be safe I stopped postfix and dovecot. I only made the mail server to learn, I don't use it as a main email or anything, but after this I'll probably move the mailboxes to a paid provider.
|
|
#
?
Aug 7, 2014 14:45
|
|
|
I almost posted this in the packrats thread, but as I'm guessing this is more of a linux question, I'll ask here instead. How can I get a list of drives not in a zfs pool? It's extremely tedious to compare the list of all disk ids in the system to the list of disk ids already in a zfs pool when I add a bunch of new drives that I want to create a new zfs pool with.
|
|
#
?
Aug 7, 2014 18:04
|
|
|
Thermopyle posted:I almost posted this in the packrats thread, but as I'm guessing this is more of a linux question, I'll ask here instead. It would probably involve listing all drives, then grepping -v in that list with the output of the list of drives that are part of the pool.
|
|
#
?
Aug 7, 2014 18:14
|
|
|
For a routed packet (i.e. coming in interface A and out interface B, not destined for a listening socket), are there any situations where iproute2 will reassemble a packet before routing it? For instance, if A has MTU 1500 and B has MTU 9000, would a 6000 byte frame ever leave B intact, or would fragments always be delivered as received?
|
|
#
?
Aug 7, 2014 23:14
|
|
|
jre posted:I'm a bit further along with this now. I have jenkins and gitorious VMs up and running and jenkins is correctly picking up pushes to the git repo. Sending every file is fine, because you're gating commits with gerrit or similar, right? So bad commits don't ever make it into the actual repo? Just unmerged patchsets? Normally I'd say "build a package, build a fresh VM with that package, that's your deployment", but I don't know how  you are. Definitely build packages and use those for deployment and versioning
|
|
#
?
Aug 7, 2014 23:19
|
|
|
NOTinuyasha posted:Ok, I'm down to this. A 5GHz 802.11n USB adapter that works with Linux, I can't find any confirmed reports of any working. Is the state of wireless on Linux really this terrible? Apologies in advance for the price. Note that I don't actually own this, but it looks kosher and is pretty specific about its Linux credentials. I did a heck of a lot of research a couple of years ago, asked around in various places and got no further than you did - this is literally the only dual band USB adapter I've found that has any evidence of Linux support. I'm sure the support is there in the kernel, but I agree it's really hard to find the right 5GHz models as a consumer. Prince John fucked around with this message at 01:20 on Aug 8, 2014 |
|
#
?
Aug 8, 2014 01:17
|
|
|
Prince John posted:Apologies in advance for the price. I use a Buffalo router as a wireless bridge to get my desktop on my home network. Sure, it ain't ideal for portable use, but at home it's nice to have the machine think it is just running ethernet. If 5ghz is important, and mobility isn't, then a cheap second router set up as a bridge would be a more cost-effective way of accomplishing getting wireless network access.
|
|
#
?
Aug 8, 2014 02:39
|
|
|
Thermopyle posted:I almost posted this in the packrats thread, but as I'm guessing this is more of a linux question, I'll ask here instead. I use this on Solaris: http://hardforum.com/showthread.php?t=1566966 I don't think there's a direct equivalent of Solaris' format command, but it might be a useful start:
|
|
#
?
Aug 8, 2014 02:48
|
|
|
evol262 posted:Normally I'd say "build a package, build a fresh VM with that package, that's your deployment", but I don't know how We're moderately  , but for this web app I don't have access to the load balancer so config changes have to be requested via email which stops me spinning up new VMs and switching over. Current deployment method is via rsync and sticky tape.It's a huge php monstrosity , so would something like this be sensible commits to testing branch get checked and sent to in house testing cluster, commits to master branch get packaged with phing and deployed via publishoverssh to remote production webservers.
|
|
#
?
Aug 8, 2014 11:12
|
|
|
edit: I guess I figured it out, thanks
the fucked around with this message at 03:33 on Aug 9, 2014 |
|
#
?
Aug 9, 2014 03:21
|
|
|
How can I boot to usb from Ubuntu? I have a formatted bootable USB for Windows 8. I went into the UEFI and "USB Bootable" or whatever is turned on. But Ubuntu seems to supercede every time I reboot and launch itself.
|
|
#
?
Aug 9, 2014 18:49
|
|
|
the posted:How can I boot to usb from Ubuntu? I have a formatted bootable USB for Windows 8. I went into the UEFI and "USB Bootable" or whatever is turned on. But Ubuntu seems to supercede every time I reboot and launch itself. If your BIOS/UEFI is set up correctly and your USB stick is set up correctly Ubuntu will have zero influence on this.
|
|
#
?
Aug 9, 2014 18:52
|
|
|
spankmeister posted:If your BIOS/UEFI is set up correctly and your USB stick is set up correctly Ubuntu will have zero influence on this. Hmm, I'm not sure what's happening then, Thanks though.
|
|
#
?
Aug 9, 2014 19:00
|
|
|
the posted:How can I boot to usb from Ubuntu? I have a formatted bootable USB for Windows 8. I went into the UEFI and "USB Bootable" or whatever is turned on. But Ubuntu seems to supercede every time I reboot and launch itself. Try another USB drive. Some don't like to boot for some reason.
|
|
#
?
Aug 9, 2014 19:05
|
|
|
Bob Morales posted:Try another USB drive. Some don't like to boot for some reason. I hope not, because I went out and bought this 16gb one because I didn't have any 
|
|
#
?
Aug 9, 2014 19:19
|
|
|
the posted:I hope not, because I went out and bought this 16gb one because I didn't have any They are like $5 at Walmart
|
|
#
?
Aug 9, 2014 21:00
|
|
|
I tried posting this here originally, but stupid me poo poo-posted it to the main forum, and the mods moved it to comedy. I've got Linux Mint 17 running in Virtualbox (works fine), and I think I've got atftpd and tacacs+ installed correctly )at least, Software manager says so. What I don't know how to do is get atftpd running (and copy CME-FULL-7.1.0.0.tar to it's root (/tftpboot, supposedly). I've got this running in GNS3 (loopback communications only, on purpose), and am trying to install CME to a C3745 ipvoice router, so I can test iphones (IP Blue MultiLab) for an upcoming cert exam. I'm a total linux newb, and have only managed to get Mint17 and Haiku running (Windows 2008R2 that I'm running PRTG and radius on WILL NOT install a tftp server, just a client) Can some kind soul please hold my hand, and walk me through this? I really need to get this working for practice, and lack a physical C375 and Cisco IP phones. I've already placed the needed tar file on the desktop. Now what do I do? I tried following someones advice for doing it in ubuntu, but I somehow hosed it up, and needed to do a fresh re-install of MINT17 (where I am now) Thank you in advance!
|
|
#
?
Aug 9, 2014 23:20
|
|
|
just installed ubuntu on an xps 15 notebook. everything runs lovely, except the fact that everything on the screen is super teeny tiny. i've tried scaling icons and window borders, but things like terminal and even default chrome scaling is just awful. any experience using ubuntu with high-dpi displays? any/all suggestions appreciated.
|
|
#
?
Aug 10, 2014 21:17
|
|
|
db franco posted:just installed ubuntu on an xps 15 notebook. Not Ubuntu specifically, but change the minimum font sizes and default zoom level for apps and browsers, which mostly fixes this. Icons may still be small depending on Windowing toolkit. Newer versions of gnome are much better.
|
|
#
?
Aug 10, 2014 21:20
|
|
|
How come the vagrant-berkshelf plugin expects my Berksfile to be in the same file as the Vagrantfile, but berkshelf expects it to be in the cookbook folder? edit: Ah, I was thinking my Vagrantfile should be in my project root (chef repo). Berkshelf generates a Vagrantfile in the cookbook directory itself with berks cookbook <name> which makes more sense when I thought about it. fletcher fucked around with this message at 01:00 on Aug 11, 2014 |
|
#
?
Aug 11, 2014 00:48
|
|
|
loose-fish posted:I assume you are using the standard xfce terminal. Just start a new terminal like this Thanks! As it turns out I had to edit the xfce4/terminal/terminalrc file to get a blinking, underline cursor rather than the block cursor, because for some reasong Xubuntu takes those options out of the GUI terminal preferences interface. But now it works and looks great.
|
|
#
?
Aug 11, 2014 11:35
|
|
|
I couldn't figure out why I would get an infinite username/password prompt trying to proxy subsonic through nginx with basic http auth. I finally fixed it by adding:code:
|
|
#
?
Aug 12, 2014 07:37
|
|
|
Don't know if there's a Fedora thread so I'll ask here. I just put my new computer together with an Asus H97-PRO motherboard with i218-v ethernet. So I figured I'd try to install Fedora since I don't have any free 64 bit Windows license available at the moment. Everything seems to work fine except the network; the LED is flashing correctly and Linux detects my hardware but the ethernet driver (e1000e) is apparently not supported by the kernel Fedora 20 uses (I think 3.11.*, I saw some forum post claiming 3.14.* should work but that doesn't really help me when I don't have a network connection). I've googled a bit but haven't really found anything that has helped me so far. The source for the driver can be downloaded and built from intel but requires the kernel headers which of course aren't available in the default installation so that's quite the catch 22. Anyone have any ideas? Or should I just give up and try another distro (Ubuntu I guess)?
|
|
#
?
Aug 12, 2014 21:37
|
|
|
e1000e is supported and has been since forever so your issue is probably something else. How about you explain what exactly the issue is you're running into.
|
|
#
?
Aug 12, 2014 21:51
|
|
|
netcat posted:Don't know if there's a Fedora thread so I'll ask here. I would suggest using nightlies. You can certainly try Ubuntu, but I think 14.04.1 is still 3.13, so it may not help. spankmeister posted:e1000e is supported and has been since forever so your issue is probably something else.
|
|
#
?
Aug 12, 2014 21:59
|
|
|
spankmeister posted:e1000e is supported and has been since forever so your issue is probably something else. My issue is that ethernet doesn't work, at all. I guess it's wrong to say that e1000e is not supported with this kernel but, yeah, what evol262 said. my hardware is detected fine so it should be a driver issue. evol262 posted:I would suggest using nightlies. You can certainly try Ubuntu, but I think 14.04.1 is still 3.13, so it may not help. Ok, I'll try the nightlies tomorrow I guess, thanks
|
|
#
?
Aug 12, 2014 22:21
|
|
|
Ok so the e1000e driver in your kernel doesn't support your chipset yet.
|
|
#
?
Aug 12, 2014 22:28
|
|
|
Place I used to work out we would use kmod packages from elrepo.org for e1000 stuff. They have one that is pretty recent, but they are based off RHEL/CentOS kernels (they do have SRPMS though): kmod-e1000e-3.1.0.2-1.el7.elrepo.x86_64.rpm 03-Aug-2014 12:02 133K http://elrepo.org/linux/elrepo/el7/x86_64/RPMS/ And here is the source for the newer module code: http://sourceforge.net/projects/e1000/files/e1000e%20stable/ Between those options and the nightly fedora, something is bound to work.
|
|
#
?
Aug 13, 2014 02:22
|
|
|
Having some more trouble with my nginx config: https://gist.github.com/fletchowns/3a7507b1c834f520ed0a When I access https://mycoolsite.com/rtgui/ the browser makes a request to https://mycoolsite.com/rtgui/submodal/common.js but it 404s: code:
|
|
#
?
Aug 13, 2014 07:28
|
|
|
Oh, hello. Apparently sometime recently a new version of Ubuntu was released... or maybe not very recently, but I have notifications only for LTS. And I'm considering opting out completely until the LTS runs out, because I have mixed experience with upgrading if my configuration is not standard. Mostly it's the Xubuntu flavor, and the fact that I've removed a bunch of standard apps. Does upgrading mess it all up, or am I misremembering or something?
|
|
#
?
Aug 13, 2014 13:15
|
|
|
How would I grep the following, if I want the first 24 hosts? Host name is as follows: hostXY01, hostXY02 ... hostXY47, hostXY48 I'm brainfarting hard, and can't come up with the regex to do it. code:
|
|
#
?
Aug 13, 2014 14:40
|
|
|
You should be fine upgrading. The real problem is the fact you're using Ubuntu and half the ends don't meet and give random loving errors for no apparent reason. I've heard of no glaring failures of 14.04 or upgrading to it; provided you're not stupidly far back on the version curve. A jump from 12.04 to 14.04 should be fairly seamless.
|
|
#
?
Aug 13, 2014 14:42
|
|
|
Anyone have any experiences, good bad or otherwise, with Oracle Linux? At first glance, it looks like another RHEL clone, only one that gets updates out faster than CentOS, and with a custom kernel (that you don't even have to use if you really want 100% RHEL compatibility). I'm looking for the downside to suggesting it over CentOS for a new project at work, and not finding it.
|
|
#
?
Aug 13, 2014 14:45
|
|
|
fletcher posted:Having some more trouble with my nginx config: https://gist.github.com/fletchowns/3a7507b1c834f520ed0a Because the reference is relative. The correct path of the file is the one nginx is looking for unless you've modified the rtgui source to use absolute paths. Weird Uncle Dave posted:Anyone have any experiences, good bad or otherwise, with Oracle Linux? At first glance, it looks like another RHEL clone, only one that gets updates out faster than CentOS, and with a custom kernel (that you don't even have to use if you really want 100% RHEL compatibility). I'm looking for the downside to suggesting it over CentOS for a new project at work, and not finding it. It's fine, except that it's Oracle. You're basically going between choosing between a known project with a long lineage (and the long delay when EL6 came out was due to project leadership issues that won't come up again) and trusting Oracle. Do you trust Oracle to continue producing OEL in the future? Are you ok not having even community support? If so, go for it. BoyBlunder posted:How would I grep the following, if I want the first 24 hosts? Or with \d hostXY([2][0-4]|[01]\d) use "grep -e" If you want to range like that, you need to use restrictive groups combined with ors.
|
|
#
?
Aug 13, 2014 15:43
|
|
|
evol262 posted:hostXY([2][0-4]|[01][0-9]) Thanks!
|
|
#
?
Aug 13, 2014 16:43
|
|
|
Weird Uncle Dave posted:Anyone have any experiences, good bad or otherwise, with Oracle Linux? At first glance, it looks like another RHEL clone, only one that gets updates out faster than CentOS, and with a custom kernel (that you don't even have to use if you really want 100% RHEL compatibility). I'm looking for the downside to suggesting it over CentOS for a new project at work, and not finding it. I've used it and it's fine. It has a few Oracle specific things that are useful if you want to run Oracle database software on it but it's nothing you can't also do with RHEL or CentOS. Question is: do you want support and if so do you want to pay Oracle for it? Also, if you're not an Oracle shop why would you choose it over RHEL or CentOS? I see no reason to use OEL if you're not an Oracle shop. And I'm not convinced their patches are quicker than CentOS, the "CentOS is slow releasing updates" hasn't been true for years. spankmeister fucked around with this message at 17:13 on Aug 13, 2014 |
|
#
?
Aug 13, 2014 17:11
|
|
|
Weird Uncle Dave posted:Anyone have any experiences, good bad or otherwise, with Oracle Linux? At first glance, it looks like another RHEL clone, only one that gets updates out faster than CentOS, and with a custom kernel (that you don't even have to use if you really want 100% RHEL compatibility). I'm looking for the downside to suggesting it over CentOS for a new project at work, and not finding it. The reason not to use it is that you would be supporting Oracle. Oracle basically rebadges RHEL and sells it for more money while calling it "Unbreakable". The reason we had to close off our knowledge base to customers only was that over 50% of our traffic was coming from Oracle call centers. The reason we had to ship all our kernel patches in one big bundle was to stop Oracle from being able to support them.
|
|
#
?
Aug 13, 2014 17:21
|
|
|
Suspicious Dish posted:The reason not to use it is that you would be supporting Oracle. Oracle basically rebadges RHEL and sells it for more money while calling it "Unbreakable". The reason we had to close off our knowledge base to customers only was that over 50% of our traffic was coming from Oracle call centers. The reason we had to ship all our kernel patches in one big bundle was to stop Oracle from being able to support them. I guess by this, a little clarification: The idea that Oracle can support Linux (and basically RHEL) better than the engineers who write it is laughable, but they tell people this, then trolled our knowledgebase looking for support articles when customers they poached called them for support. Patches come in a bundle available separately so Oracle actually has to backport their own stuff instead of trying to tweak around patches for backported fixes in the kernel SRPM. Incidentally, this is also why their "Unbreakable" kernel is no longer compatible -- they'd actually have to put in engineering effort to get it working on the EL kernel and shipping upstream is easier for them.
|
|
#
?
Aug 13, 2014 17:32
|
|
|
|
| # ? Apr 29, 2024 18:44 |
|
|
Geeze. I know almost nothing about Oracle, but everything I ever hear about them makes them sound like real shitheels.
|
|
#
?
Aug 13, 2014 18:33
|
|