Installation error when adding user to vboxsf group #189
Comments
|
Interesting, I haven't seen this before yet, but you reproduced it on two distributions. In any case, I implemented a (temporary) patch that discards this error so it's not a show-stopper anymore. |
|
Hi Maarten, Thanks for getting back to me. I'm sure I entered the correct sudo password ;-) There is only one normal user account on my installations and it has sudo rights. I attempted installation in virtual machines (virtualbox), then selected 1) in a local user environment, used my home folder for the installation, and chose the stable version. With your patch I get past the vboxsf error. btw: I see you install gensim and friends. Perhaps I asked this before, but will you be adding its lsa/lda implementation to LaMachine? Groet, |
|
Hi, I'm running into a new problem here ;-) Lamachine is started with: Alpino, both woprs, frog, foliadocserve etc. processes are all started. Accessing the server through http://hostname, http://127.0.0.1 or http://localhost makes no difference. When I use the default tscan settings, it starts analysis but then shortly after reports "error obtaining status". Log from tscan says this: |
|
The woprs reloading continuously sounds very much like a memory issue. How much memory do you have available in the virtual machine? Wopr is very memory-hungry, you'll need like 5 to 6GB per wopr, and then there's Frog and Alpino which may grow in memory when used, so don't attempt it all on a VM with anything less than 16GB. |
|
Thanks, yes I know these are eating memory. I tried with 16 and 24 Gb. |
|
That could be an issue with the foliadocserve entry in the portal yes, if FLAT works fine than the document server is working as it should. Regarding wopr, there may be some further clues in /usr/local/var/log/uwsgi/tscan.uwsgi.log . If you have enough memory then I wouldn't expect the continuous restarts you experience. |
|
This is the tscan log after a clean boot and start of lamachine. Both woprs seem to be listening: *** Starting uWSGI 2.0.19.1 (64bit) on [Wed Nov 25 13:18:51 2020] ***
Size of InstanceBase = 13640943 Nodes, (545637720 bytes), 30.21 % compression Size of InstanceBase = 14249860 Nodes, (569994400 bytes), 30.18 % compression |
|
and the uwsgi log: |
|
Ok, I reinstalled lamachine and now run it with 24 Gb. The woprs made system load go up to 20Gb memory but were stable when analysing a test file. So it was a memory issue, but the foliaserver issue is still present. Best, |
|
But the foliadocserve issue only shows when you access it through the portal right? FLAT works fine? That may be a bug I'll have to look into then. |
|
Hi, yes that is the case. |
Hi Maarten,
I attempted installation of LaMachine on both Ubuntu 20 and Fedora 33.
Both fail with:
TASK [lamachine-core : Adding user to vboxsf group for shared folder access (if it exists)] ***************************************
fatal: [localhost]: FAILED! => {
"changed": false,
"name": "xxx", (username anonymised)
"rc": 10
}
MSG:
usermod: Permission denied.
usermod: cannot lock /etc/group; try again later.
Anything I can do to prevent this?
The text was updated successfully, but these errors were encountered: