00:31 | vagrantc has left IRC (vagrantc!~vagrant@2600:3c01:e000:21:7:77:0:20, Quit: leaving) | |
01:03 | wyre is now away: Auto away at Fri Feb 17 01:02:56 2023 UTC | |
03:20 | jgee1186922 has left IRC (jgee1186922!~jgee@186.80.49.20, Quit: The Lounge - https://thelounge.chat) | |
03:25 | jgee11869225 has joined IRC (jgee11869225!~jgee@186.80.49.20) | |
04:13 | <alkisg> !tftp
| |
04:13 | <ltspbot> tftp: To test if your TFTP server is working, try this on the server (or from another PC): busybox tftp -g -r /ltsp/ltsp.ipxe -l /tmp/ltsp.ipxe localhost
| |
04:16 | vagrantc has joined IRC (vagrantc!~vagrant@2600:3c01:e000:21:7:77:0:20) | |
04:54 | vagrantc has left IRC (vagrantc!~vagrant@2600:3c01:e000:21:7:77:0:20, Quit: leaving) | |
06:23 | wyre is back | |
06:38 | <vsuojanen[m]> hi, do you remember if is there is any discussion or documentation how to make <100mbps clients nfs homedirectory connection faster? I have just two specific clients with same specs and speed, the clients have issues with nfs autologin when they boot, it seems autologin happens too fast when RELOGIN_TIMEOUT=5. sometimes even the home is not mounted. it works stable when RELOGIN_TIMEOUT=15
| |
06:45 | wyre is now away: Auto away at Fri Feb 17 06:45:11 2023 UTC | |
06:46 | <vsuojanen[m]> i know the problems is the old slow cable that needs fix also but same desk works with other client hardware
| |
06:54 | <alkisg> vsuojanen: I'm not sure what you're asking, if a connection is <100 mbps due to a cabling problem, how would you make it faster using software?
| |
06:58 | <vsuojanen[m]> not quite that, I remember two or three years ago there was some discussion about nfsroot settings and if i run now /usr/lib/klibc/bin/nfsmount from the client when they mount homedir does it need something with slow connections
| |
06:59 | something extra
| |
07:02 | <quinox> you want to reduce the initial mount time, not throughput?
| |
07:02 | <vsuojanen[m]> yes
| |
07:04 | <quinox> I have no direct ideas, but you could do `man 5 nfs` and then try out any option that sounds useful
| |
07:05 | I would suggest fixing the cabling though if possible :)
| |
07:05 | (does the bad cables also cause packet loss?)
| |
07:05 | <vsuojanen[m]> i want to have the cable fixed of course sure
| |
07:08 | yes, some busy hours in the day. I don't know why but users say it's hanging nightly :)
| |
07:11 | <quinox> if the ping stays consistent but you got packet loss I would _reduce_ timeout values and _increase_ retry attempts. If the pick fluctuates and you got packet loss then ... ¯\_(ツ)_/¯
| |
07:11 | s/pick/ping/
| |
07:12 | I see something called "soft timeout" that causes increased responsiveness with an added risk of data corruption
| |
07:14 | I would be hesitant to even go this route: nothing as permanent as temporary workarounds. If people feel the pain of the bad cables right now this gives you more leverage to get it fixed, I'd say
| |
07:14 | <vsuojanen[m]> thanks so sum up if I really want there are some settings i could try with /usr/lib/klibc/bin/nfsmount options?
| |
07:15 | it has not been any better without ltsp clients so that's why i call it stable client with RELOGIN_TIMEOUT=15
| |
07:17 | <alkisg> vsuojanen: solving bad packet handling in all the related linux software may take 100 years. Solving the cabling issue may take 10 minutes. Choose where to spend your effort wisely :)
| |
07:17 | wyre is back | |
07:17 | <alkisg> A couple of years ago I was investigating nfs speed issues. NOT bad packet handling for NFS.
| |
07:17 | It's a completely different topic
| |
07:18 | The speed issues were related to buffering and caching
| |
07:18 | <vsuojanen[m]> thanks. I'm only thinking possible shortcuts so not too bad yet
| |
08:19 | woernie has joined IRC (woernie!~werner@p200300c56f3a82009e43a16456bf1736.dip0.t-ipconnect.de) | |
08:28 | ricotz has joined IRC (ricotz!~ricotz@ubuntu/member/ricotz) | |
08:31 | ricotz is now away: I'm busy | |
13:03 | <vsuojanen[m]> quinox: alkisg Have a nice weekend
| |
13:06 | wyre is now away: Auto away at Fri Feb 17 13:05:41 2023 UTC | |
13:43 | woernie has left IRC (woernie!~werner@p200300c56f3a82009e43a16456bf1736.dip0.t-ipconnect.de, Remote host closed the connection) | |
13:44 | Vercas has left IRC (Vercas!~Vercas@gateway/tor-sasl/vercas, Remote host closed the connection) | |
13:44 | Vercas has joined IRC (Vercas!~Vercas@gateway/tor-sasl/vercas) | |
13:50 | <quinox> you too!
| |
13:50 | Vercas has left IRC (Vercas!~Vercas@gateway/tor-sasl/vercas, Ping timeout: 255 seconds) | |
13:51 | Vercas has joined IRC (Vercas!~Vercas@gateway/tor-sasl/vercas) | |
14:19 | Vercas5 has joined IRC (Vercas5!~Vercas@gateway/tor-sasl/vercas) | |
14:21 | Vercas has left IRC (Vercas!~Vercas@gateway/tor-sasl/vercas, Ping timeout: 255 seconds) | |
14:21 | Vercas5 is now known as Vercas | |
14:52 | wyre is back | |
14:57 | woernie has joined IRC (woernie!~werner@p5ddec58c.dip0.t-ipconnect.de) | |
15:01 | <alkisg> You too :)
| |
15:06 | lcurl has left IRC (lcurl!~UserNick@115.79.27.26, Read error: Connection reset by peer) | |
15:07 | lcurl has joined IRC (lcurl!~UserNick@115.79.27.26) | |
16:07 | wyre is now away: Auto away at Fri Feb 17 16:07:20 2023 UTC | |
17:00 | jgee118692253 has joined IRC (jgee118692253!~jgee@186.80.49.20) | |
17:02 | jgee11869225 has left IRC (jgee11869225!~jgee@186.80.49.20, Ping timeout: 246 seconds) | |
17:02 | jgee118692253 is now known as jgee11869225 | |
17:02 | lcurl has left IRC (lcurl!~UserNick@115.79.27.26, Read error: Connection reset by peer) | |
17:03 | lcurl has joined IRC (lcurl!~UserNick@115.79.27.26) | |
17:19 | lcurl has left IRC (lcurl!~UserNick@115.79.27.26, Read error: Connection reset by peer) | |
17:19 | lcurl has joined IRC (lcurl!~UserNick@115.79.27.26) | |
18:47 | woernie has left IRC (woernie!~werner@p5ddec58c.dip0.t-ipconnect.de, Remote host closed the connection) | |
18:59 | vagrantc has joined IRC (vagrantc!~vagrant@2600:3c01:e000:21:7:77:0:20) | |
20:06 | wyre is back | |
23:22 | ricotz has left IRC (ricotz!~ricotz@ubuntu/member/ricotz, Quit: Leaving) | |