ViciBox v.8.0.1 Released

[Kumba]

It's the stanard .Oops release, aka v.8.0.0.

I figured it reached the point where it became necessary to shoot the engineers and begin production so here it is:
http://download.vicidial.com/iso/vicibo ... -8.0.0.iso

95% of the ViciBox v.7.0 manual follows the V.8.0.0 install so doing the install documentation is going to take a few days. Here is what comes in ViciBox v.8

- OpenSuSE Leap v.42.3 64-bit
- Linux Kernel v.4.4.87
- Asterisk v.11.25.1-vici with support/upgrade to Asterisk v.13.12.2-vici
- DAHDI v.2.11.1
- LibPRI v.1.6.0
- VoiceSync v.1.3.8
- MariaDB v.10.1.22
- Apache2 v.2.4.23
- CertBot v.0.15.0 (Free LetsEncrypt SSL Certificates)
- ViciDial SVN Trunk v.2.12-533a build 160106-1342 revision 2461


This is the first release using the 'kiwi' image building system so I am sure there will be some things here and there that will need to be improved. Please let me know of any bugs or installation issues you run into. Some notable things are that it defaults to standard text based interfaces upon booting. The only graphical boot sequence is Grub after it's installed and I just need to figure out how to get that fixed in a reliable fashion. Also gone is that network bug where you had to double-boot it to get the network to initialize properly.

[covarrubiasgg]

root# bless you !!!

Upgrading some servers right away....

[elondoninc]

Wahoo! Boy am I glad I cleared my cache on my browser today, and noticed that it went to 8.0. Athankyou

[mike.mc]

Super !!! Merciii

[Kumba]

Just rolled all the bug fixes into v.8.0.1. Feel free to download it and give it a try.

[proper]

Thank you much, appreciate your hard work.

[TwistedFister]

can you roll into v8 straight from v7? what is the upgrade process consist of? I am currently at v7 2.14-579a SVN 2641

[frequency]

SCP wont work, reported regarding kex-alg, fixed that sshd_config

Can you please tell where the mysql files now reside? I am unable to find in /var/lib/mysql

EDIT: seems to be in /srv/mysql/data

chown -R wwwrun:root /var/spool/asterisk/monitorDONE
Permission issues

Thanks

[Kumba]

MySQL data directory was move to /srv/mysql/data in order to prepare the server for having separate SSD drives. The binary log and regular logging are in their normal spot, but the actual database resides under /srv/mysql/data. You would mount your SSD partition as /srv/mysql to get the proper effect. This was done to support more boutique SSD drives like NVMe which don't always boot correctly. Also allows you to use a relatively inexpensive drive for the OS and spend money on a high-quality SSD that doesn't need to be so big.

Regarding scp, are you saying a particular key algorithm you use for SSH is not loaded by default? Looks like it's more of a legacy SSH issue. I'll put it into the bugfix notes and add it to 8.0.2 when I roll that out. Not sure how old of a Linux install you would need for that to be an issue but it'll be in there. If you can send me what you had to change I'll put that in he bugfix. Otherwise I have nothing that generates this error so I can't really give any instructions to people.

I'll look into the permision issue and add it to the bugfix thread once I verify it.

[frequency]

KexAlgorithms diffie-hellman-group1-sha1

We added this in sshd_config. When we tried to scp the DB from a centOS server onto vicibox8, we added this in this config.

Tried using the manual we received from Vicidial support and still unable to use the webphone successfully . Phone is still "connecting". no firewall on server, hardware firewall disabled.

Certicates are in pem format and in /etc/letsencrypr/live/fqdn/ folder. Does the location or format type make a difference?

MySQL data directory was move to /srv/mysql/data in order to prepare the server for having separate SSD drives. The binary log and regular logging are in their normal spot, but the actual database resides under /srv/mysql/data. You would mount your SSD partition as /srv/mysql to get the proper effect. This was done to support more boutique SSD drives like NVMe which don't always boot correctly. Also allows you to use a relatively inexpensive drive for the OS and spend money on a high-quality SSD that doesn't need to be so big.

Regarding scp, are you saying a particular key algorithm you use for SSH is not loaded by default? Looks like it's more of a legacy SSH issue. I'll put it into the bugfix notes and add it to 8.0.2 when I roll that out. Not sure how old of a Linux install you would need for that to be an issue but it'll be in there. If you can send me what you had to change I'll put that in he bugfix. Otherwise I have nothing that generates this error so I can't really give any instructions to people.

I'll look into the permision issue and add it to the bugfix thread once I verify it.

[Kumba]

Can you go to https://fqdn and get a proper SSL Connection? If so then your certbot/letsencrypt/apache config is correct.

As far as the web phone that depends on which one you are using.

KexAlgorithms diffie-hellman-group1-sha1

We added this in sshd_config. When we tried to scp the DB from a centOS server onto vicibox8, we added this in this config.

Tried using the manual we received from Vicidial support and still unable to use the webphone successfully . Phone is still "connecting". no firewall on server, hardware firewall disabled.

Certicates are in pem format and in /etc/letsencrypr/live/fqdn/ folder. Does the location or format type make a difference?

MySQL data directory was move to /srv/mysql/data in order to prepare the server for having separate SSD drives. The binary log and regular logging are in their normal spot, but the actual database resides under /srv/mysql/data. You would mount your SSD partition as /srv/mysql to get the proper effect. This was done to support more boutique SSD drives like NVMe which don't always boot correctly. Also allows you to use a relatively inexpensive drive for the OS and spend money on a high-quality SSD that doesn't need to be so big.

Regarding scp, are you saying a particular key algorithm you use for SSH is not loaded by default? Looks like it's more of a legacy SSH issue. I'll put it into the bugfix notes and add it to 8.0.2 when I roll that out. Not sure how old of a Linux install you would need for that to be an issue but it'll be in there. If you can send me what you had to change I'll put that in he bugfix. Otherwise I have nothing that generates this error so I can't really give any instructions to people.

I'll look into the permision issue and add it to the bugfix thread once I verify it.

[frequency]

ntp daemon is not added by default..causing time sync issues and asterisk crash at times (not sure about asterisk though).

[Kumba]

I just downloaded and installed ViciBox v.8.0.1 from the website. NTP is running by default, so I am not sure what happened for it to not be running on your system.

ntp daemon is not added by default..causing time sync issues and asterisk crash at times (not sure about asterisk though).

[frequency]

I just downloaded and installed ViciBox v.8.0.1 from the website. NTP is running by default, so I am not sure what happened for it to not be running on your system.

ntp daemon is not added by default..causing time sync issues and asterisk crash at times (not sure about asterisk though).

i think i will have to check this by reinstalling because my asterisk is crashing with 20 inbound calls with 30 agents with load average going as high as 11.00 when it crashes. very strange. never witnessed that before on any vici7 and yes we are using the new webrtc phone as well.

Code: Select all

[Oct 11 16:44:33] WARNING[9665]: chan_sip.c:4039 retrans_pkt: Retransmission timeout reached on transmission 376184150-223829631-2133773997 for seqno 2 (Critical Response) -- See https://wiki.asterisk.org/wiki/display/AST/SIP+Retransmissions
Packet timed out after 31999ms with no response
[Oct 11 16:44:41] WARNING[9665]: chan_sip.c:4039 retrans_pkt: Retransmission timeout reached on transmission 1234756390-152418122-840864098 for seqno 2 (Critical Response) -- See https://wiki.asterisk.org/wiki/display/AST/SIP+Retransmissions

[chornyi_taras]

Are you using webrtc phone?

[williamconley]

i think i will have to check this by reinstalling because my asterisk is crashing with 20 inbound calls with 30 agents with load average going as high as 11.00 when it crashes. very strange. never witnessed that before on any vici7 and yes we are using the new webrtc phone as well.

Pay attention to the Inbound agi scripts, they are complex and can be challenging and may be reacting differently for some reason on the new platform. If they have a memory leak, that can be detrimental even after the call has been routed and the agi should have dropped.

[chornyi_taras]

I just downloaded and installed ViciBox v.8.0.1 from the website. NTP is running by default, so I am not sure what happened for it to not be running on your system.

ntp daemon is not added by default..causing time sync issues and asterisk crash at times (not sure about asterisk though).

i think i will have to check this by reinstalling because my asterisk is crashing with 20 inbound calls with 30 agents with load average going as high as 11.00 when it crashes. very strange. never witnessed that before on any vici7 and yes we are using the new webrtc phone as well.

Code: Select all

[Oct 11 16:44:33] WARNING[9665]: chan_sip.c:4039 retrans_pkt: Retransmission timeout reached on transmission 376184150-223829631-2133773997 for seqno 2 (Critical Response) -- See https://wiki.asterisk.org/wiki/display/AST/SIP+Retransmissions
Packet timed out after 31999ms with no response
[Oct 11 16:44:41] WARNING[9665]: chan_sip.c:4039 retrans_pkt: Retransmission timeout reached on transmission 1234756390-152418122-840864098 for seqno 2 (Critical Response) -- See https://wiki.asterisk.org/wiki/display/AST/SIP+Retransmissions

this is known asterisk 11 issue https://issues.asterisk.org/jira/plugin ... RISK-25275. I have a good news for you. Patch that fixes this issue is already included in vicibox repos. http://vicidial.org/VICIDIALforum/viewt ... 63#p130263

[frequency]

this is known asterisk 11 issue https://issues.asterisk.org/jira/plugin ... RISK-25275. I have a good news for you. Patch that fixes this issue is already included in vicibox repos. viewtopic.php?f=8&t=37706&p=130263#p130263

Thank you and yes, using the new WebRTC phone and also tried your PBXWebPhone. Both replicating the same issue.

i think i will have to check this by reinstalling because my asterisk is crashing with 20 inbound calls with 30 agents with load average going as high as 11.00 when it crashes. very strange. never witnessed that before on any vici7 and yes we are using the new webrtc phone as well.

Pay attention to the Inbound agi scripts, they are complex and can be challenging and may be reacting differently for some reason on the new platform. If they have a memory leak, that can be detrimental even after the call has been routed and the agi should have dropped.

Dialplans and agi's seems to be working with 2 other closer servers.

[chornyi_taras]

this is known asterisk 11 issue https://issues.asterisk.org/jira/plugin ... RISK-25275. I have a good news for you. Patch that fixes this issue is already included in vicibox repos. viewtopic.php?f=8&t=37706&p=130263#p130263

Thank you and yes, using the new WebRTC phone and also tried your PBXWebPhone. Both replicating the same issue.

i think i will have to check this by reinstalling because my asterisk is crashing with 20 inbound calls with 30 agents with load average going as high as 11.00 when it crashes. very strange. never witnessed that before on any vici7 and yes we are using the new webrtc phone as well.

Pay attention to the Inbound agi scripts, they are complex and can be challenging and may be reacting differently for some reason on the new platform. If they have a memory leak, that can be detrimental even after the call has been routed and the agi should have dropped.

Dialplans and agi's seems to be working with 2 other closer servers.

This is becase viciphone and PbxWebPhone are based on same js library. Did you upgrade asterisk to latest version? It should resolve this issue with crashing

[williamconley]

Dialplans and agi's seems to be working with 2 other closer servers.

Note that the AGI script for the Ingroup runs on the server that receives the call. It then passes the call to an agent wherever it finds one, but that script doesn't run on the other servers.

If this server is "the one" that receives the inbound DIDs, then this script is suspect. If it's not, then ... not so much.

PS: This is why we suggest that rooms with primarily inbound calls should rotate the calls between servers, to avoid this script overloading. It's usually not a problem if there is never a waiting queue for calls, but if there is a backup ... whoa.

[frequency]

Code: Select all

Configuring Telephony Server...
/usr/bin/sox FAIL formats: can't open input file `./*.gsm': No such file or directory
Loading GMT and Phone Codes...

Is this normal error?

[uselessinfoguru]

I had the same issue with 8.0.1, ntp daemon was not installed with the os-install command, i had to run yast timezone, then install the ntp daemon from there after the os-install, before the vicibox-express install

[Kumba]

I just downloaded ViciBox v.8.0.1 from our website and NTP daemon was running. I'll double check it again, it's possible I missed something.


****

I just installed ViciBox v.8.0.1 after downloading it from vicibox.com. After os-install I typed 'ntpq' and then typed 'o at the CLI and I see the time servers it's synchronizing too. I also see ntpd running by typing 'service ntpd status'.

So, I am not sure how or why NTP isn't starting on your installs. I'm unable to duplicate that issue. What does the output of 'service ntpd status' say after installing before you start it or modify anything?

[uselessinfoguru]

i'll let you know after i reinstall, i'm restoring a backup from vicibox 7.0.3 right now to see if everything works right

[uselessinfoguru]

something went wrong with the restore, not sure what, but now i get this from welcome.php
MySQL connect ERROR: Access denied for user 'cron'@'localhost' (using password: YES)
and from asterisk -r
Unable to connect to remote asterisk (does /run/asterisk/asterisk.ctl exist?)
restarting the 8.0.1 install now

[uselessinfoguru]

first thing i notice is during the os-install, after accepting the license agreement, when setting date and time, it will not allow me to select ntp, the accept button does not work with ntp selected, i must set it to manual time...
updates in a moment

Code: Select all

Dialer:~ # service ntpd status
● ntpd.service - NTP Server Daemon
   Loaded: loaded (/usr/lib/systemd/system/ntpd.service; enabled; vendor preset: disabled)
  Drop-In: /run/systemd/generator/ntpd.service.d
           └─50-insserv.conf-$time.conf
   Active: active (running) since Wed 2017-10-18 11:30:32 EDT; 5min ago
     Docs: man:ntpd(1)
 Main PID: 1691 (ntpd)
    Tasks: 2 (limit: 512)
   CGroup: /system.slice/ntpd.service
           ├─1691 /usr/sbin/ntpd -p /var/run/ntp/ntpd.pid -g -u ntp:ntp -c /etc/ntp.conf
           └─1716 ntpd: asynchronous dns resolver                               

Oct 18 11:30:31 vicibox8 systemd[1]: Starting NTP Server Daemon...
Oct 18 11:30:32 vicibox8 ntpd[1673]: ntpd 4.2.8p10@1.3728-o Fri Jul  7 17:41:17 UTC 2017 (1): Starting
Oct 18 11:30:32 vicibox8 ntpd[1673]: Command line: /usr/sbin/ntpd -p /var/run/ntp/ntpd.pid -g -u ntp:ntp -c /etc/ntp.conf
Oct 18 11:30:32 vicibox8 ntpd[1691]: proto: precision = 1.160 usec (-20)
Oct 18 11:30:32 vicibox8 ntpd[1691]: switching logging to file /var/log/ntp
Oct 18 11:30:32 vicibox8 systemd[1]: Started NTP Server Daemon.
Oct 18 11:30:32 vicibox8 start-ntpd[1632]: Starting network time protocol daemon (NTPD)


so the service is running, but when i do yast timezone, and set to ntp i get

These packages need to be installed: xx
xxyast2-ntp-client

then after package install i get
Files /etc/ntp.conf have been changed manually.x
xYaST might lose some of the changes

[williamconley]

when setting date and time, it will not allow me to select ntp, the accept button does not work with ntp selected

Probably because it's the presently selected configuration and does not represent a change. Therefor "save" is not needed.

And if you have ntp active, "setting" the time manually is not supported (that's what NTP is ... time is set by inter-server communication, not a dude at the console).

Just a guess.

[uselessinfoguru]

during os-install when i go to other settings, manual is selected by default

[williamconley]

during os-install when i go to other settings, manual is selected by default

Which could be an indication of what they expect you to select.

Code: Select all

service ntpd status

If this result is "active", then ntp is active.

Code: Select all

ntpq -p

If this result shows '*' or '+' in the far left column, you are synced with that remote server.

The rest is interesting and open to interpretation and must be translated through "linux engineers are farkin' weird" before deciding what they mean and what their goal is. None of which matters: Either ntp is active and synced ... or not.

[uselessinfoguru]

well, fresh install, same thing, ntp didn't sync until i ran yast timezone, and it auto-installed yast2-ntp-client
on the plus side the restore from a backup up vicibox 7.0.3 worked wonderfully on the 2nd install attempt (still not sure what happened the first time), everything seems good to go here now..
for others, if you can't get an ntp sync during os-install, use manual time, then after os-install run yast timezone, and set it to ntp there, it will then install the additional repositories required

spoke too soon, after a minute or so i'm still getting this after the restore
MySQL connect ERROR: Access denied for user 'cron'@'localhost' (using password: YES)

[williamconley]

well, fresh install, same thing, ntp didn't sync until i ran yast timezone, and it auto-installed yast2-ntp-client
on the plus side the restore from a backup up vicibox 7.0.3 worked wonderfully on the 2nd install attempt (still not sure what happened the first time), everything seems good to go here now..
for others, if you can't get an ntp sync during os-install, use manual time, then after os-install run yast timezone, and set it to ntp there, it will then install the additional repositories required

NTP will not sync if time is off by too much. In that case, it's necessary to turn off ntp, set the time manually or sync manually, and then turn ntp back on. There's a threshold of how much ntp is allowed to adjust. If it's off by too much, it simply refuses to sync.

That's not a function of the installer, it's a question of the hardware clock and often an indication that the CMOS battery is dying.

[uselessinfoguru]

good point, Will, just so happens this testing pc has a low voltage issue with the battery, i.e. battery is dying

still don't know what the issue is with the database tho, after restoring the backup everything comes up fine, carriers connect, phones register, then all of a sudden, asterisk unloads, i can no longer access the system from the web, i can still ftp or ssh in, asterisk is no longer running tho

trying to install from trunk i get this

Code: Select all

DBI connect('asterisk:localhost:3306','cron',...) failed: Access denied for user 'cron'@'localhost' (using password: YES) at ./install.pl line 2689.
Couldn't connect to database: Access denied for user 'cron'@'localhost' (using password: YES) at ./install.pl line 2689.


[williamconley]

Restore is not a simplistic task. If you restored configuration files, especially, you probably broke the system since configuration files for asterisk, apache and mysql are not "upgrade proof". (In fact, I doubt configuration files for any package are upgrade-proof)

Vicidial scripts and web pages are stable across versions. In fact, that is the purpose of the Vicibox installer iso: It creates a stable environment for all the vicidial scripts. But each time OpenSuSE changes versions of any underlying software, those packages potentially change their configuration methods. Implanting an old config file on top of the new one is ... unlikely to be a healthy choice.

[uselessinfoguru]

ok, so, either go with vicibox 7.0.4 or... is there a way to restore the asterisk database and vici admin without messing with the configuration files?

[williamconley]

When you follow the restore procedures, you decide what to restore. Don't mess with package configuration files.

[uselessinfoguru]

so, if i restore just the etc folder and the database to mysql... that 'might' work?

[williamconley]

I'm not sure if you put this together or not, but: the etc folder is all "configuration files". Considering that we just discussed avoiding restoring configuration files ... you may want to rethink that statement.

Just sayin'

(PS: Or to be more clear: Your goal is to restore the DB and the Web files and then modify the server under Admin->Servers to be sure that those values are updated to reflect the present server's new characteristics, such as asterisk version and external IP address. You'll also need to run the IP Update Script if the new server's IP is not the same as the previous one.).

[uselessinfoguru]

ok so, how do the phones and users get restored to the admin? wouldn't i need to restore the backup sip.conf files and extensions.conf files?

[williamconley]

sip.conf and extensions.conf do not contain information specific to your install.

sip-vicidial.conf and extensions-vicidial.conf do. However: Those files are generated by Vicidial!

Let me ask you a question: When you create a new phone in Admin->Phones, do you then also have to edit a configuration file? (I'll answer that for you: NO.)

If you have manually edited any files in /etc/, there is likely to be something unusual going on that quite possibly doesn't need to go on at all with one exception: externip in sip.conf. This value is often manually modified when you change your external IP if you are using NAT. But that value would be properly created in your freshly installed Vicibox 8.0.1 server and not need to be modified during a restore. That value is only changed if the public IP of the server changes.

[uselessinfoguru]

so, if i'm understanding you correct, all i need to do is restore the database using the mysql command, then copy (restore from backup) the srv folder, then use admin to rebuild files, then reboot (just to insure everything loads properly) and, if needed, edit the externip line in sip.conf