Hi, after restart Turris v.1.1. web administration is not accessible, only SSH.
(there is now message: web 192.168.1.1 refuse connection)
I haven’t done any changes for long time.
I tried restart lighttpd:
/etc/init.d/lighttpd restart
/usr/sbin/lighttpd: can’t load library ‘libpcre.so.0’
Everything else is working (NAS, web).
Any idea?
Thanks
After last OS update , i am facing similar issue.
Luci and Basic web-admin pages are not accessible.
Restarting lighttpd fix that for a while. I will check logs later on. Maybe some config entry is dummy now.
On my iPhone I was forced to reset my cookies and other web settings to be able to login via safari.
I had no problem logging in after update on my mac using safari.
Try to reinstall libpcre package.
Did update finished correctly? There should be both ne pcre, but also new lighttpd, so both of them should be updated. which should fix your issue.
When I open web administration after OS update, I have to use username and password. I tried root admin etc…but it didn’t work. When I try restart lighttpd I got this:
root@turris:/# /etc/init.d/lighttpd restart
Syntax OK
/usr/lib/lighttpd/mod_alias.so
/usr/lib/lighttpd/mod_cgi.so
2016-12-09 11:54:14: (plugin.c.187) Cannot load plugin mod_auth more than once, please fix your config (lighttpd may not accept such configs in future releases)
2016-12-09 11:54:14: (plugin.c.227) dlopen() failed for: /usr/lib/lighttpd/mod_authn_file.so File not found
2016-12-09 11:54:14: (server.c.912) loading plugins finally failed
Hi,
I had the same error message “can’t load library 'libpcre.so.0” when I was trying to use grep (full version, not from busybox). Because of that other services also failed, like unbound and postfix. Updating grep solved it.
Could you write me how to do it, google didn’t help me 
Thanks
Try opkg update; opkg install libpcre.
Also try running the updater.sh -n again.
SOLVED:
I used
opkg upgrade
And its now working 
Thanks
Luci and Basic web-admin pages are not accessible by me too - only sometime and a short time (< 20 sec). I either wait or I tried also https and then there worked.
Hi Tomskra,
I’m facing similar issue and started extra topic about it:
Thanks @JOhnyRi, I simply restarted Lighthtpd via Terminal and LuCI was back on
root@Arcadia:~# /etc/init.d/lighttpd restart
Syntax OK
/usr/lib/lighttpd/mod_alias.so
/usr/lib/lighttpd/mod_cgi.soSimilar failure here 
Since the recent (2016/12/16) automatic system update the web admin interface does not work anymore.
As I was able to use SSH at least I found the lighthttpd refuses to load screaming it requires a rrd binary to be set. Why did not the updater take care of that? What the … … … … … … (lots of 4-letter words), what’s wrong with collecting some statistics enabled by built-in means???
I’m not a openwrt/Linux guru, I used LuCi (now unavailable) to confugure the box.
Btw, my MC was broken by the update as well – furtunately deleting its (not too precious) history helped (as found here, thanks).
Any idea other than RTFM (i.e. spending a year reading The Linux Bible and trying zillions of googled-out links which may or may not work)? TIA…
Being a T-O owner does not necessarily qualify one as an expert in this area. Contrary, additionally to performing its primary networking-related tasks the T-O represents a neat platform for learning and playing which may eventually bring the expertise some day.
H.
PS: Dear SW maintainers, are you learning from a well known company which releases regular upgrades often breaking things that used to work? Don’t do that, please.
MC was and will not be broken by any update. The bug you refer to existed even before and apparently has been reported.
Were you able the restart lighttpd now?
Thank you for your effort.
Yet, according to the referred post (which I used to cure my instance) it seems MC failure having something common with updates – NONES’s last post in the original thread in Czech in the old forum (“…I should take care of MC update myself…”)
Nevertheless, it is not the key issue.
No, that’s why I asked the community for help. To make it clear I approximately did the following:
2016-12-18 02:43:59: (log.c.217) server started
2016-12-18 02:43:59: (mod_rrdtool.c.391) rrdtool.binary has to be set
2016-12-18 02:43:59: (server.c.1270) Configuration of plugins failed. Going down.
Where are the days of NetWare 3.12/4.x which could run for months without a reboot (even if it was patched several times during its up period) and with patches that never brought it to its knees? (At least this is my personal experinece when I was baby-sitting a few of them – I usually almost didn’t know we had them.)
H.
Additional steps:
H.
I have not touched the collection or rrd stuff in Omnia so I don’t know the correct configuration for it but the error message indicates mod_rrdtool plugin for lighttpd (probably coming from package lighttpd-mod-rrdtool) doesn’t find its configuration.
You can try to remove lighttpd-mod-rrdtool package in case you don’t need it or if you have reinstalled it and the web admin still doesn’t work you could try something like this as a workaround:
echo 'rrdtool.binary = "/usr/bin/rrdtool"' > /etc/lighttpd/conf.d/local.conf
/etc/init.d/lighttpd restart
Thanks white,
Unfortunately
2016-12-18 11:06:27: (server.c.1296) WARNING: unknown config-key: rrdtool.binary (ignored)
Removing the new local,conf.
Nevertheless, I found some info on lighttpd config in the meantime and renamed the rrd load which brought my LuCi to life again,
Now I’ve been trying to reinstall the rrd/collecd stuff (must be removed first even if it was uninstalled through opkg). I guess I have to delve into /etc/config/updater and other conf docs stuff anyways.
Have a nice and really sunny Sunday (no Sun here shining outside neither down on my Turris).
H.
My mistake. You are right. Sorry. 