[02:09] U1 has been acting up for the past week or so. It gives this error repeatedly http://paste.ubuntu.com/443740/ . I noticed the bulletin from May 24th, but I had to seek help. The sync daemon is using 400-500 MB of ram. [02:22] hello u1helpplease [02:22] yes hello [02:23] u1helpplease: they will be testing the new fixes tommorrow, so hopefully we will be seing improvements in performance before the end of the week [02:24] thanks for your time. Can the sync daemon be run with ulimit -m? [02:27] u1helpplease: hmm, i am not sure [02:45] it doesn't i tried "ulimit -m 64000 -v 128000 /usr/lib/ubuntuone-client/ubuntuone-syncdaemon" no luck still 500mb thanks though. Best of luck tomorrow. I will have another 25 gig coming your way. [02:47] u1helpplease: if you come by tommorrow between 12:00-20:00 GMT roughly, a lot of the devs are on [03:56] you would need to put the ulimit on python probably === kermiac is now known as kermiac_work === kermiac_work is now known as Guest42358 === Guest42358 is now known as kermiac_working === cpg is now known as cpg|biab === cpg|biab is now known as cpg [09:28] whats the 10.04 procedure to sync the files in 2 dirs on 2 diff machines, it seems the folders are not syncing to the server? [09:30] koshari, our servers are thrashing like crazy [09:30] we're working on scaling them [09:30] but we're still having our ups and downs [09:30] i see [09:33] beuno because a new kernel has just hit the repos? [09:34] koshari, no, because we grew 200k users in 2 months :) [09:36] beuno: ah one of the pitfalls of success i guess :-) [10:21] honk. i'm about to lose patience with U1. how's this upgrade of the server infrastructure coming along? [10:22] diverse_izzue, we're working non-stop [10:22] it gets increasingly harder, because people's syncs are backed up [10:22] so we add more capacity and tune performance, but load continues to increase [10:23] we have more work being rolled out that should help us get over the hump [10:23] beuno, ok, sorry for bitching. i can stop running it for a while, and try again in a week or so [10:23] I've been trying to sync for a few days as well :( [10:23] diverse_izzue, it's ok, we have almost 500k users, so a few won't make a difference [10:24] leave it on, we hope it will be back to normal-ish in a few days [10:24] but, beuno, a general question. i enabled u1 sync on the directory where my phd thesis lives that i'm working on, because i want that backed up. of course those files are very regularly changed and saved. is that a problem for u1 (in a scenario where the servers are running fine)? [10:25] diverse_izzue, it's not a problem, no [10:25] at the moment, we don't upload deltas, so you will upload the full file each time you save [10:25] it's not optimal, but it won't create problems [10:26] beuno, i am actually running into problems because it often creates .u1conflict files and then my latex compilation fails [10:26] diverse_izzue, are you working on different machines? [10:27] if its on the same machine, then it's absolultely a bug [10:27] beuno, it is only one machine [10:27] diverse_izzue, it should not happen [10:27] i have u1 enabled on 2 machines, but the other machine is not online now [10:28] diverse_izzue, edit .config/ubuntuone/syncdaemon.conf [10:28] and add: [10:28] [logging] [10:28] level = DEBUG [10:28] restart syncdaemon [10:28] the next time you get a u1conflict [10:29] wait i don't understand [10:29] should i change the file now? [10:29] file a bug report and include the .cache/ubuntuone/log/syncdaemon.log [10:29] yes [10:29] so you can log more stuff [10:29] and then when it happens restart the daemon? [10:29] it will start logging in more detail [10:30] so when you hit this problem again [10:30] if you hand us the ~/.cache/ubuntuone/log/syncdaemon.log file [10:30] we can find the problem [10:30] ok, i changed logging level now, will let you know if it happens. thanks for the help. i'll let you work now and do sth. myself. bb. [10:31] diverse_izzue, np, thanks === cpg is now known as cpg|away [13:03] good morning! [13:09] duanedesign, morning! [13:15] hey duanedesign, rye :) [13:17] hey duanedesign [13:23] rye: I was talking to bodhizazen and I mentioned your latest blog post. He wanted me to pass along this link. He thought you might find it helpful. http://www.redhat.com/docs/en-US/Red_Hat_Enterprise_Linux/6-Beta/html/Resource_Management_Guide/pr02.html [13:25] duanedesign, cgroups... nomnomnom - i thought it was lxc-specific... [13:25] He said "Those pages are fast reading and review cgroups, best linky I know of for cgroups, which of course is how one manages LXC containers / processes ...." === kenvandine_ is now known as kenvandine [15:03] Current state - service is being reconfigured to run through additional layer of database pool. As you might already know extremely high number of unnecessary database connections slowed the system down [16:23] hi [16:23] why my Ubuntu One folder doesn't update? :( === dendrobates is now known as destro === destro is now known as dendrobates [16:33] sylarpowa, currently the file sync is in process of being updated to become a lot faster [16:34] sylarpowa, you might have noticed that during last 4 weeks performance was gradually decreased and today first 2 servers were switched to new system [16:35] rye: actually I think no file has been updated, is there any way to check the status of the update on my system? [16:36] sylarpowa, you can run u1sdtool --waiting-meta in the terminal to see the items that are queued in meta queue - https://wiki.ubuntu.com/RomanYepishev/UbuntuOne/ClientControl [16:40] rye: thanks [16:42] sylarpowa, there will be an announcement on identi.ca/mailing-list once we find out that file sync performance is restored [16:43] rye: actually I'm trying to update some files from many days, maybe 2 weeks [16:43] but nothing happens [16:50] sylarpowa, that's exactly the symptoms i am describing. It should be possible to perform full sync later today [16:52] * mkarnicki has his fingers crossed, needs it badly for gsoc === teknico is now known as teknico_away [18:41] Evening all. I purchased some music this morning on via the U1 Music Store and it is still marked as queued and has yet to be moved to my U1 storage....is there a known issue? [18:42] honk [18:49] PaulGit: you can wait for a better reassuring answer, but I can tell you there is *much* maintanance done today over the server side [18:49] PaulGit: so I think it's much related. [18:50] mkarnicki: Thanks for that. I have just looked via the webui and the files are there so the Rhythmbox UI is telling lies....need to look into why they are not being downloaded to my desktop.... === JUMPA is now known as jumpa [19:00] PaulGit: huh.. I see. still, it's sync related, maybe it's not your machine, maybe it's the servers. good luck. [19:01] mkarnicki: I will leave it and see what happens, I have plenty of other music to listen too! :) [19:03] =) === cpg|away is now known as cpg [19:20] hi fecundo and rye and others [19:20] two days ago I wrote about 2.6.34-5 kernel and ubuntuone [19:21] because 2.6.34-5 kernel seems break U1 [19:21] (desktopcouch really) [19:22] Are there news about this issue? [19:24] gio, there is a bug report for that, let me find it for you... [19:24] thank you [19:26] gio, bug #588478 [19:26] Launchpad bug 588478 in desktopcouch (Ubuntu) (and 1 other project) "RuntimeError: Can not find port of couchdb. (affects: 8) (dups: 3) (heat: 54)" [High,Confirmed] https://launchpad.net/bugs/588478 [19:28] thank you. In your opinion, is this a kernel related bug or a desktopcouch bug? [19:31] I do this question because I'm using 2.6.34-4 and U1 works fine [19:34] gio, i feel that's dc bug, since parsing of /proc/$pid/net/tcp does not produce any results. Either it should search better or do something better to determine listening port [19:36] but, strangely, with 2.6.34-4 dc works, so the question is: why? :) [19:39] gio, hmmm, in 2.6.32 /proc/$pid/net/tcp contains more entries than in .34 in maverick [19:40] Well, I try a comparison later [19:40] thanks [19:40] bye === jumpa is now known as JUMPA === yofel_ is now known as yofel [21:50] ping one.ubuntu.com responds, but viewing http://one.ubuntu.com not [21:50] ^ [21:50] btw [21:51] 504 Gateway Time-out [21:51] just a quick note - we are currently looking into the issue that ... yes, mkarnicki exactly! [21:51] aha :) [21:51] mkarnicki, replication was enabled for couchdb that has caused some strange effects, the team is checking what has happened [21:52] rye: cool, thank you for explanations :) [21:53] rye: you're a good soul, always around helping others. good job man [22:05] rye: heh, good - I've learned my application should handle such situations also ;) [22:05] mkarnicki, right now something is really strange happening with couchdb [22:06] rye: o_O [22:06] sounds scary === cpg is now known as cpg|away [22:24] So Ubuntu One seems to have forgotten my computer; it wants me to add my computer to my account now