| === Lcawte is now known as Lcawte|Away | ||
| === zz_DenBeiren is now known as DenBeiren | ||
| === peter is now known as Guest65138 | ||
| === markthomas is now known as markthomas|away | ||
| === Guest58311 is now known as Locke2002 | ||
| === Lcawte|Away is now known as Lcawte | ||
| fivetwentysix | The russians are abusing my postfix server | 11:07 |
|---|---|---|
| fivetwentysix | YES! I win! | 11:11 |
| fivetwentysix | postqueue -p | 11:11 |
| fivetwentysix | Mail queue is empty | 11:11 |
| fivetwentysix | Down from 23430424 emails | 11:12 |
| fivetwentysix | lol | 11:12 |
| Sachiru | LOL | 11:24 |
| Sachiru | GJ | 11:24 |
| === Lcawte is now known as Lcawte|Away | ||
| === kickinz1|afk is now known as kickinz1 | ||
| === kickinz1 is now known as kickinz1|afk | ||
| === kickinz1|afk is now known as kickinz1 | ||
| === kickinz1 is now known as kickinz1|afk | ||
| === chmurifree is now known as chmuri | ||
| Valduare | hmm | 16:28 |
| Valduare | if I make an image of a drive will it do just the used space or the entire drive partition size | 16:29 |
| qman__ | imaging takes an exact bit copy of everything, including free space | 16:29 |
| qman__ | imaging is rarely if ever required and usually not the best choice | 16:29 |
| Valduare | k ty | 16:30 |
| === markthomas|away is now known as markthomas | ||
| === markthomas is now known as markthomas|away | ||
| Valduare | qman__: I got a windows laptop here thats got a failing hard drive | 16:58 |
| Valduare | still readable data | 16:59 |
| Valduare | takes half hour to boot to windows so I was thinking of taking a copy of the hd before going further | 16:59 |
| qman__ | Valduare: you should use gnu ddrescue | 17:00 |
| cfhowlett | Valduare, stop booting. it's failing with each read/write. get thee to a backup solution immediately | 17:00 |
| qman__ | get a new hard drive to copy it to, or take an image if you have to, but use gnu ddrescue to make a recovery image | 17:00 |
| Valduare | im booted into ubuntu live cd atm | 17:00 |
| Valduare | havnt heard of gnu ddrescue | 17:00 |
| Valduare | looking it up now | 17:00 |
| Valduare | is it a direct 1to1 solution | 17:01 |
| Valduare | ie I need to have the new hd available | 17:01 |
| qman__ | the short of it: ddrescue -f -n /dev/sda /dev/sdb /tmp/rescue.log | 17:01 |
| qman__ | you should, that's the best way to do it | 17:01 |
| qman__ | but you don't have to, you can do to a file if required | 17:01 |
| qman__ | it won't be compressed though | 17:02 |
| qman__ | and it will be an exact 1:1 copy of everything it was capable of recoverin | 17:02 |
| Valduare | just the used space right? | 17:03 |
| qman__ | all the space, total size of the disk | 17:03 |
| Valduare | prob with these huge 1 terabyte hds lol | 17:03 |
| Valduare | damn | 17:03 |
| Valduare | so what if im wanting to go from a 1 terabyte platter drive to a 256 ssd | 17:04 |
| qman__ | can't do it | 17:04 |
| Valduare | there’s only 71 gigs used space on the 1 terabyte drive | 17:04 |
| cfhowlett | qman__, as you're clearly better informed than I ... I buy a new computer. I'm going to do some experimental things. I want to image the drive in case I need to restore it to virginal status. How to proceed? does the backup include the empty space on the drive? | 17:04 |
| qman__ | the recovery drive must be as large or larger than the failing drive | 17:04 |
| qman__ | once you have recoered the data, you can then move it to a smaller drive | 17:04 |
| qman__ | the recovery tools are filesystem agnostic, because they have to be | 17:05 |
| qman__ | so they cannot recognize what matters and what doesn't | 17:05 |
| qman__ | you should turn off the system with the failing drive and wait until you can get a same size or larger drive to recover to | 17:06 |
| === Lcawte|Away is now known as Lcawte | ||
| qman__ | cfhowlett: imaging is not a good backup solution, file backups are more space efficient, faster to take, faster to recover, and avoid fragmentation | 17:07 |
| Valduare | its fine to have it booted up into ubuntu live cd | 17:07 |
| Valduare | hd isnt accessed | 17:07 |
| qman__ | no, it isn't | 17:07 |
| qman__ | as long as that drive is spinning, it could die completely | 17:08 |
| qman__ | shut down, unplug the drive | 17:08 |
| Valduare | we’re not completely sure “what kind of dieing” the hd is | 17:08 |
| cfhowlett | Valduare, yep. this ^^^ | 17:08 |
| qman__ | it doesn't matter what kind | 17:08 |
| qman__ | cut power until you can recover | 17:08 |
| qman__ | that's the safest bet, always | 17:08 |
| cfhowlett | Valduare, assume the worst kind, assume the worst time and proceed from there. | 17:08 |
| === bilde2910 is now known as bilde2910|away | ||
| qman__ | cfhowlett: imaging is best for forensics, and as you see here, recovering from a failing hard drive | 17:11 |
| cfhowlett | qman__, fair enough. I'll research more. my goal was to restore "reset" to pre-me mucking about status. | 17:12 |
| qman__ | cfhowlett: an image will be a 100% bit-for-bit copy, including fragemntation, noise in your free space, your empty free space, and all | 17:12 |
| Valduare | read error rate 136 | 17:12 |
| qman__ | Valduare: stop messing with it, shut down, and unplug the disk until such time as you have another 1TB or larger disk to recover to; every second you spend with it spinning increases chances of total failure | 17:13 |
| Valduare | its doing a compressed backup to an external hd atm | 17:13 |
| qman__ | cfhowlett: if it has to be _exactly_ the same, an image is the way to go, but be aware of the downsides; most of the time, things don't actually need to be exactly the same | 17:17 |
| cfhowlett | qman__, thank you much. | 17:17 |
| Valduare | looks like 1 terabyte platter drive 85 at best buy | 17:31 |
| Valduare | guess i’ll go that route instead of the ssd for their laptop heh | 17:31 |
| RoyK | gnu ddrescue rocks | 17:44 |
| RoyK | I once got a drive from a user that couldn't be mounted on any OS i tried it on | 17:44 |
| RoyK | it took gnu ddrescue a little less than a week to recover 99,8% of the data | 17:45 |
| cfhowlett | RoyK, 24 x 7 for a week? hardcore! | 17:47 |
| RoyK | 500GB | 17:47 |
| RoyK | 24x5.5 or thereabouts | 17:48 |
| RoyK | (24x7 for a week doesn't make sense - a week for a week :P) | 17:48 |
| === Xbert is now known as Guest1097 | ||
| qman__ | yeah, I had one that kept turning off | 17:52 |
| qman__ | but if I power cycled it, it would spin back up | 17:52 |
| qman__ | repeated that for about a week and a half, got well over 99% of the data, enough to recover | 17:52 |
| qman__ | less than a megabyte was lost | 17:53 |
| === DenBeiren is now known as zz_DenBeiren | ||
| === markthomas|away is now known as markthomas | ||
| === bilde2910|away is now known as bilde2910 | ||
| === bilde2910 is now known as bilde2910|away | ||
| blaaa | I have just restarted a server remotely after a firmware update, but the network is not coming up. I have no terminal unfortunately.... the only thing I can see is the network interfaces are actually coming up at post, but they are shut down at some point while ubuntu is booting | 21:15 |
| === esde is now known as Guest18963 | ||
| bekks | you need some terminal or console access to investigate that issue. | 21:16 |
| === Guest18963 is now known as esde | ||
| blaaa | bekks: that's probably true, but some hypothesis on what could have caused this would be nice... before I'll be on my way. | 21:18 |
| blaaa | bekks: I think maybe the issue is caused by renamed interfaces after the fw update | 21:18 |
| blaaa | i think NIC interface names might not be persistent if some enumeration has changed, is that right? | 21:19 |
| qman__ | blaaa: the interface names are based on MAC address | 21:29 |
| qman__ | blaaa: a boot failure is much more likely - by default, if it fails to boot once, grub will wait indefinitely the second time around | 21:30 |
| qman__ | blaaa: an fsck or failure to mount something could also be at issue | 21:30 |
| blaaa | qman__: thanks, i'll have to go and see I suppose. After the firmware update the computer booted fine (but I had not tested the network, not even attached it....) and it did a clean acpi shutdown after a complete boot | 21:39 |
| RoyK | qman__: do you know how to configure grub to stop that nonesense? | 21:43 |
| RoyK | nonsense, even | 21:44 |
| blaaa | RoyK: there is some option in the config | 21:44 |
| blaaa | RoyK: don;t remember what it is right now though | 21:44 |
| blaaa | had looked it up when setting up the server... | 21:45 |
| qman__ | RoyK: I have, I looked it up before | 21:46 |
| qman__ | It was causing me issues with some ebayed supermicro servers that rarely boot the first time | 21:46 |
| qman__ | http://askubuntu.com/questions/178091/how-to-disable-grubs-menu-from-showing-up-after-failed-boot | 21:46 |
| blaaa | was it GRUB_RECORDFAIL_TIMEOUT ? | 21:46 |
| qman__ | yes | 21:47 |
| blaaa | yes, apparently | 21:47 |
| RoyK | qman__: thanks | 21:47 |
| blaaa | lucky the server is not far away... | 22:16 |
| blaaa | apparently the interface name had changed | 22:17 |
| blaaa | from p1p2 to eth0 | 22:17 |
| blaaa | p2p1... | 22:17 |
| blaaa | but I have no idea why, except for a fw update and reboot-cycle nothing had changed | 22:18 |
| blaaa | and I believed the point of the new naming scheme was to have persistent names... | 22:18 |
| === bilde2910|away is now known as bilde2910 | ||
| furkan | i've got a question about grub: i'm running 14.04 in a VM, and i modify /boot/grub/menu.lst to add a virtual serial port so that i can console in with virsh, and whenever there's a new kernel installed it likes to update menu.lst and overwrite my 2 lines - any way i can avoid this? | 22:32 |
| furkan | these are simply the 2 lines that i put in: | 22:32 |
| JanC | furkan: sre | 22:32 |
| furkan | serial --unit=0 --speed=115200 --word=8 --parity=no --stop=1 | 22:32 |
| furkan | terminal --timeout=10 serial console | 22:33 |
| JanC | sure | 22:33 |
| furkan | JanC: i'm all ears :) | 22:33 |
| furkan | i noticed the /etc/grub.d/40_custom file but wasn't sure if that was the right place to put it since it said custom menu entries | 22:34 |
| JanC | it's explained in the manual AFAIK, but you need to put it in a script in /etc/grub.d/ | 22:34 |
| JanC | those scripts get run as part of update-grub | 22:35 |
| furkan | so when i run update-grub is it supposed to get copied into menu.lst? | 22:36 |
| furkan | since i just tried putting it there and it doesn't seem to do that | 22:36 |
| JanC | the output of those scripts gets copied to menu.lst | 22:36 |
| JanC | so yes, you need to run update-grub | 22:37 |
| furkan | man i'm confused, on my desktop 14.04 i've got grub2 but i guess the server version is still using the old one | 22:38 |
| JanC | oh | 22:38 |
| JanC | you sure? | 22:38 |
| JanC | Amazon? | 22:38 |
| furkan | no well this used to be a 12.04 VM | 22:38 |
| furkan | and then i upgraded to 14.04 | 22:38 |
| furkan | is grub2 default in ubuntu server 14.04? | 22:38 |
| JanC | grub2 was the default in 10.04 and before | 22:39 |
| furkan | hmm maybe because i used vmbuilder | 22:39 |
| === bilde2910 is now known as bilde2910|away | ||
| JanC | but IIRC there is/was some issue with Amazon cloud images not working with that (never needed that) | 22:40 |
| furkan | ya this is on our own server, not amazon | 22:40 |
| furkan | we've got a bare metal running 14.04 + a few 14.04 VMs running some services | 22:41 |
| JanC | it's still possible that it is/was used in cloud images though? | 22:41 |
| furkan | i built the image with vmbuilder for 12.04... so whatever that does by default i guess | 22:42 |
| furkan | but yeah i guess this is sort of odd since /etc/grub.d seems to be for grub2 only | 22:43 |
| JanC | that's right | 22:43 |
| furkan | i wonder if anything will break if i upgrade | 22:44 |
| furkan | so for grub2 it looks like the right place to put the serial console stuff is in /etc/default/grub | 22:47 |
| furkan | unless that's the same thing as just putting it in the custom file | 22:47 |
| JanC | not the same thing | 22:49 |
| JanC | the variables you set in /etc/default/grub are used by the scripts in /etc/grub.d/ | 22:49 |
| furkan | ah i see, interesting | 22:50 |
| furkan | thanks for the info :) | 22:50 |
| JanC | it's interesting to read those scripts in /etc/grub.d/ some time :) | 22:53 |
| === markthomas is now known as markthomas|away | ||
| === bilde2910|away is now known as bilde2910 | ||
| === bilde2910 is now known as bilde2910|away | ||
| === Lcawte is now known as Lcawte|Away | ||
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!