[05:06:58] What? [05:10:22] Gloria: ?? [05:10:58] Betacommand: [22:28:14] Cyberpower678, Gloria: where do I should file a bug for the by you maintained https://tools.wmflabs.org/erwin85/ ? [05:12:00] (03PS1) 10Tim Landscheidt: Simplify toolwatcher [labs/toollabs] - 10https://gerrit.wikimedia.org/r/122094 [05:52:10] A medium size instance has 40G HDD storage. If I run df-h I get the following output physikerwelt@math2:/etc/mysql$ df -h Filesystem Size Used Avail Use% Mounted on /dev/vda1 7.6G 1.9G 5.3G 26% / udev 2.0G 8.0K 2.0G 1% /dev tmpfs 792M 260K 791M [05:52:31] For me it's not obvious where the mysql data-directory should go to? [09:20:17] Hello! I wanted to ask about one tool that I want to run on Labs: it looks through the recent IP's that edited Wikipedia and checks whether they are open proxies by scanning ports and trying to connect to them: if proxy works for bot, it blocks IP on Wikipedia. I haven't found prohibition of such tasks in "Terms of Use" but wanted to consult whether it is acceptable for Labs. [09:51:47] Hi. I get "Warning ... the server's host key does not match one PuTTY has..." Has anyone made changes? [09:58:48] MGA73: yes [09:59:09] MGA73: servers have changed all of bastions were reinstalled [09:59:31] Super... Thank you :-) [12:18:10] Coren: around [14:50:34] MGA73: if you're not already, best to subscribe to https://lists.wikimedia.org/mailman/listinfo/labs-l so you hear about things like that in advance :) [15:25:32] liangent: Kinda. I'm trying to get a sorta day off, but if there's an emergency I'm around. [18:13:54] !log tools Removed /data/project/backup: Only empty dynamic-proxy backup files of January 3rd and earlier [18:13:58] Logged the message, Master [18:16:14] !log tools Removed empty directories /data/project/{d930913,sudo-test{,-2},testbug{,2,3}}: Corresponding service groups don't exist (anymore) [18:16:17] Logged the message, Master [18:31:48] Hi! I already have putty public and private key. But now I want to open a session running on another computer. I have public key, username and passphrase. But I have not the private key. How can I get access anyway? [18:45:30] Is it possible? [18:56:39] is anybody here? [18:56:39] Hi Doc_Taxon, I am here, if you need anything, please ask, otherwise no one is going to help you... Thank you [18:56:53] I did [18:56:59] Hi! I already have putty public and private key. But now I want to open a session running on another computer. I have public key, username and passphrase. But I have not the private key. How can I get access anyway? [19:02:54] Doc_Taxon: you cant [19:03:06] okay [19:03:21] you need to have your private key to connect [19:03:39] okay [19:03:42] thank you [19:03:53] Its a security issue [19:29:39] Doc_Taxon: yep. you need a private key. but you can generate a fresh pair on your other computer and upload the public key [19:30:48] ya okay, question solved. Thank you very much. [19:33:48] Doc_Taxon: yw [19:34:19] scfc_de: is there a timeout / autologout on -dev? [19:35:07] my connection to labs used to be quite fast but now it is very laggy, hm … [19:36:21] command was idle for ~2h while performing a sql transaction [19:36:47] and this autologout killed it :/ [19:37:02] gifti: mine too [19:37:23] hedonil: There shouldn'T be any timeout or auto-logout. [19:37:38] hedonil: Are you sure it wasn't killed for OOM? [19:38:57] scfc_de: pretty sure. as said was idle while performing an sql command [19:40:50] failure [19:40:58] !failure [19:40:59] Cyberpower678 is responsible! [19:41:04] hehe [19:42:38] hedonil: I don't see anything killed in /var/log/syslog, so you may be right. My assumption is based on that if you use the mysql command line client without --quick, it can use a lot of memory. Do you remember/still have the error message of ssh? [19:44:31] scfc_de: this happened with other connectios [19:44:51] scfc_de: before - even on -login [19:45:21] scfc_de: now (since ~2days or so) -login is stable, while -dev isn't [19:51:10] terminal multiplexer for persistence? [19:51:19] hedonil: I lose connections every time my DSL modem reconnects :-). If you want to play it safe, you can either run your query as a grid job or use screen (I'm not sure if mosh provides the same functionality). On tools-dev, your last log entry in auth.log is: "14:33:31: pam_unix(sshd:session): session closed for user hedonil", and that's the only entry related to that sshd pid. [19:53:57] scfc_de: I usually rely on screen or better stored procedure call, but this time it hit me [19:55:03] but you can try to reproduce it by yourself with a blank open connection on -dev. and wait for it to be terminated [19:55:34] beyound your DSL reconnect ;) [20:04:27] Judging from the intervals between when my IRC nick changes to scfc_de`, I /cannot/ reproduce that :-). I don't remember being kicked out without DSL reconnects or OOM/reboots, and I'm definitely not in the mood for debugging phantom pains :-). [20:04:41] hedonil: Do you remember/still have the error message of ssh? [20:06:39] scfc_de: it's more than a phantom pains as it's real. but the message was simply like session closed when I returned. [20:08:05] scfc_de: started a blank session at 8 pm UTC, see what will happen [20:10:47] hedonil: What ssh client are you using? [20:14:14] scfc_de: OpenSSH client / Linux [20:16:22] scfc_de: but as said, behavior is different on -login and -dev since ~2 days. -login holds connection, -dev closes it if idle [20:20:45] hedonil: I only find "session closed" in a comment and in sftp-server.c. Are you sure that was the error message? [20:21:27] scfc_de: afaik there was no further error message, just plain: closed [20:39:02] Is someone else getting "No webservice" a lot? [20:47:10] Josve05a: you need to finish migration and restart webservice [23:40:23] scfc_de: With your connection issue I think I had a similar case. [23:41:04] After the move to eqiad I had to enable keep alive pings in my ssh client otherwise any period of inactivity would close my session