| 00:45.33 | jfindlay | ~napoleon |
| 00:45.33 | infobot | There's like a buttload of gangs at this school. One of them kept wanting me to join because I'm pretty good with a bo-staff. |
| 00:47.16 | jfindlay | keldwud: cassette or reel? |
| 01:28.42 | *** join/#utah infobot (ibot@208.53.50.136) |
| 01:28.42 | *** topic/#utah is Utah-related discussion including user groups | logs @ http://infobot.rikers.org/%23utah/ | See also ##lds | To combat spam, you must register with NickServ if you wish to speak. This may or may not be temporary. |
| 01:28.42 | *** mode/#utah [+o infobot] by ChanServ |
| 02:10.22 | JPaul | keldwud, what changer/drive are you using that you got for less than $2k? |
| 02:12.02 | JPaul | I'd love to replace the scalar 100 we're using, thing is so old and it keeps having trouble with the picker |
| 02:12.40 | klys | ~slashdot |
| 03:07.46 | jlp_znc | JPaul: I think he meant the tapes were less than $2k each. :D |
| 03:08.29 | JPaul | that'd make more sense |
| 03:09.00 | JPaul | last time I checked just one drive was a few grand or something like that. |
| 03:10.54 | jlp_znc | he said it was a quantum superloader 3 changer and he was using lto7 tapes. |
| 03:11.35 | jlp_znc | used: https://www.amazon.com/Quantum-SuperLoader-3-Tape-Autoloader/dp/B00B332A60 |
| 03:12.40 | jlp_znc | under $1k (just the changer). here's a 20 pack of tapes for $1245: https://www.amazon.com/Pack-HP-C7977A-LTO-7-Tape-6-25/dp/B01JN817EM/ref=sr_1_3?s=electronics&ie=UTF8&qid=1543979517&sr=1-3&keywords=LTO7+tape |
| 03:13.46 | jlp_znc | and here's a drive for $2800 https://www.amazon.com/Quantum-Drive-Components-Other-TC-L72AN-BR/dp/B0188OO370/ref=sr_1_2?s=electronics&ie=UTF8&qid=1543979591&sr=1-2&keywords=LTO-7+tape+drive |
| 03:16.36 | jlp_znc | I've got a couple of DAT drives and a stack of tapes kicking around somewhere, but it would take me a while to find them. |
| 03:22.24 | JPaul | hmmm |
| 03:23.14 | JPaul | I'd have to compare the throughput to our current setup. right now we run 6 drives in a scalar 100 to permit concurrent jobs. we have about 68 different backup jobs that run each weekend onto tape |
| 03:23.36 | JPaul | but it's onto lto3... |
| 03:26.28 | *** join/#utah emcnabb (~emcnabb@50.225.197.140) |
| 04:30.35 | *** join/#utah nagefire_ (~nagefire@66-219-236-168.dynamic.ip.veracitynetworks.com) |
| 04:46.57 | jlp_znc | oh, wow, lto3... that's pretty old. still, it works. |
| 04:55.14 | *** join/#utah tiwula (~lane@174-23-66-242.slkc.qwest.net) |
| 11:19.07 | *** join/#utah thin_ (~vvvvv@unaffiliated/thin) |
| 13:03.57 | *** join/#utah emcnabb (emcnabb@nat/redhat/x-ihjfcqtssgdfkqpf) |
| 13:51.11 | *** join/#utah aeryndunham (~aeryndunh@96.71.45.243) |
| 13:55.20 | *** join/#utah aeryndunham_ (~aeryndunh@96.71.45.252) |
| 15:07.35 | *** join/#utah josephscott (~josephsco@74.114.153.251) |
| 15:19.28 | *** join/#utah nagefire_ (~nagefire@66-219-236-168.dynamic.ip.veracitynetworks.com) |
| 15:22.02 | keldwud | jfindlay: cassette |
| 15:23.00 | keldwud | JPaul: using the quantum superloader 3 autochanger. holds two magazines and 16 tapes and has one drive. not sure how much it is, but it looks like jlp_znc might have linked it |
| 15:23.57 | keldwud | JPaul: I'd be happy to share my experiences, though. over the last week out of frustration I stopped taking copious notes, however. |
| 15:24.48 | keldwud | I just did an 8 TB backup job, the average throughput was about 190MB/s |
| 15:25.08 | JPaul | over gigabit ethernet? |
| 15:25.20 | keldwud | over SAS |
| 15:25.30 | keldwud | I have the device directly connected to my backup server |
| 15:26.22 | JPaul | how is your backup server configured? is it just treated like a NAS or does it grab backups from other servers |
| 15:26.41 | keldwud | this is the first device we've implemented in our environment so I havent set it up yet to run backup jobs from other clients |
| 15:26.47 | JPaul | gotcha |
| 15:26.50 | keldwud | but bareos/bacula is capable of it |
| 15:27.17 | keldwud | and since the local backups are directed htrough the client anyway, whatever I've set up locally would translate easily to other clients |
| 15:27.32 | JPaul | yeah, we use bacula and run two servers with it, one is the main bacula server, the other is the controller for the scalar and SCSI interface to the drives |
| 15:27.38 | keldwud | but yeah that's how it would work, it would grab backups from other server |
| 15:28.01 | keldwud | I also wouldn't recommend the superloader if you need to run multiple jobs concurrently |
| 15:28.21 | keldwud | I mean, between bacula and the superloader 3, it can do multiplexing |
| 15:28.24 | JPaul | every morning it runs the 68 jobs we have right now that reach out to all the client servers and pull down backups and either store them on a shelf for the daily incrementals, |
| 15:28.33 | keldwud | but it sounds like you need at least two drives minimum |
| 15:28.38 | JPaul | or onto tape for the weekly differentials and monthly full's |
| 15:28.42 | keldwud | and the autochanger I have only has the one drive |
| 15:29.15 | keldwud | how much data are you backing up? |
| 15:30.20 | JPaul | not a lot really, the incremental's staty relatively small, the differentials take between 7-9 LTO3 tapes, and the monthly around 20-25 LTO3 tapes |
| 15:30.31 | JPaul | *stay |
| 15:31.07 | keldwud | hell, all you really need to do is just upgrade the drive so you can use better tapes :) |
| 15:31.12 | JPaul | i'd be nice to at least go to LTO5 or something, but it's the cost and time commitment to upgrade that's holding us back |
| 15:31.34 | JPaul | if I could get the two jr admins we're looking for here I could eventually have the time |
| 15:32.39 | JPaul | I need a new scalar as well though, the picker in ours keeps having issues and causing the backups to fail over the weekend |
| 15:32.50 | JPaul | it's pretty dang old |
| 15:33.45 | JPaul | I'm pretty mechanically and electrically inclined, so if I had the time to spend on it I could probably refurbish the picker. But it's that time thing again. :-) |
| 15:34.30 | JPaul | It's only Wednesday and I'm already almost 24 hours in this week |
| 15:36.22 | JPaul | first thing though is I need to move the backup vlan off the core network and onto a dedicated physical network. half of it is already off thanks to an emergency we had at the beginning of the year |
| 15:36.35 | JPaul | just need to move the other half |
| 15:59.16 | keldwud | hmm... strategies for verifying my 8TB backup |
| 15:59.27 | keldwud | I tested a 245MB backup and restore successfully |
| 15:59.40 | keldwud | but I failed to add the "verify" check in my 8TB job |
| 15:59.51 | keldwud | and I don't have 8TB of free space available to test a restore |
| 16:05.59 | JPaul | checksums? |
| 16:06.19 | JPaul | I don't like backup testing, despite it's wisdom |
| 16:06.26 | keldwud | I dunno how to checksum the data on the tapes |
| 16:06.37 | keldwud | since it's not a filesystem |
| 16:06.41 | JPaul | not sure if it's possible, just throwing it out there |
| 16:06.46 | keldwud | :) |
| 16:07.06 | JPaul | I'd like backup testing if I had someone to pawn it off onto.. |
| 16:07.14 | keldwud | yeah I was thinking of maybe installing a bareos client on another machine with 12TB free then waiting 11 hours then running a checksum there |
| 16:07.35 | keldwud | like, I just want to trust the backup so I can delete the original file to free up space |
| 16:07.40 | keldwud | but I know better |
| 16:08.26 | JPaul | I know that if you're ever doing a restore on something that is just at the retention period cutoff, do NOT let anyone restart bacula... |
| 16:25.37 | *** join/#utah aeryndunham (~aeryndunh@96.71.45.252) |
| 16:56.01 | *** join/#utah frostyfrog (~frostyfro@unaffiliated/frostyfrog) |
| 17:29.58 | *** join/#utah braxtron (~braxtron@216.160.248.178) |
| 17:31.46 | *** join/#utah josephscott (~josephsco@74.114.153.251) |
| 17:37.49 | *** join/#utah tiwula (~lane@70.89.246.241-Busname-UT.hfc.comcastbusiness.net) |
| 17:51.32 | *** join/#utah aeryndunham (~aeryndunh@96.71.45.252) |
| 17:54.14 | *** join/#utah yukaia (sid110254@gateway/web/irccloud.com/x-bpymsvypvibzibkh) |
| 17:56.09 | *** join/#utah SunSparc (~jonathan@45.56.57.85) |
| 18:42.51 | *** join/#utah josephscott (~josephsco@74.114.153.251) |
| 18:43.22 | *** join/#utah aeryndunham (~aeryndunh@96.71.45.252) |
| 18:58.11 | *** join/#utah moccos (~moccos@104.243.62.132) |
| 19:02.56 | *** join/#utah lane_ (~lane@70.89.246.241-Busname-UT.hfc.comcastbusiness.net) |
| 19:03.07 | *** join/#utah BLZbubba (~mark@185.56.20.124) |
| 19:07.40 | *** join/#utah lane__ (~lane@70.89.246.241-Busname-UT.hfc.comcastbusiness.net) |
| 22:03.09 | jlp_znc | ~middleman |
| 22:03.10 | infobot | Lord love a duck! |
| 22:19.02 | keldwud | I read that wrong.... |
| 22:19.41 | jlp_znc | joshh: you were at Uber, weren't you? |
| 22:21.31 | joshh | ya |
| 22:21.56 | jlp_znc | I have a recruiter looking for SRE candidates for them... as I recall you didn't have a lot of good things to say. |
| 22:22.13 | joshh | hmm really? don't remember that exactly |
| 22:22.24 | joshh | but i think its probably ok, and better now than in a while |
| 22:22.26 | jlp_znc | (I'm not interested in relocating myself, but wonder if it would be something I could recommend to my contacts). |
| 22:22.46 | jlp_znc | yeah? maybe I'm confusing you with someone else. |
| 22:22.58 | joshh | i might have talked about some problems at one point |
| 22:23.11 | joshh | but i think it's fine to recommend to people overall |
| 22:23.20 | jlp_znc | okay... good to know. thanks! |
| 22:23.45 | joshh | np :) |
| 22:24.34 | joshh | a fair amount of pretty interesting stuff to do and see actually, i like it |