Even fresher install of 4.6

[Kurt Albershardt]

Built another Centos 6 VM from scratch this afternoon:

Allocated 3 CPUs and 4GB of RAM to the VM
ran 'yum update' and restarted VM.
Competed 4.6 install per wiki.
ran sipxecs-setup, answered questions, got:
Configuring system, this may take a few minutes...
then
for the next ~50 minutes, memory usage on the VM stayed at 2.76GB and CPU at 52-53% (of 3 CPUs.)
then memory jumped to 3.99GB, java restarted several times with lower CPU and RAM numbers.
after about 5 minutes, sipxecs-setup ended with the expected:
done.

Now I see a similar java CPU load as the previous VM install had:

5084 sipx 20 0 3014m 546m 13m S 227.0 13.3 9:07.83 java

[***@sipx ~]# ps aux -ww | grep 5084
sipx 5084 211 13.3 3086832 559556 ? Ssl 17:41 11:44 /usr/bin/java -Dcom.ibm.tools.attach.enable=no -Dprocname=sipxconfig -XX:MaxPermSize=128M -Xmx1024m -Djava.io.tmpdir=/var/sipxdata/tmp -Djetty.lib.dir=/usr/share/java/sipXecs/sipXconfig -Djetty.conf.dir=/etc/sipxpbx -Djetty.log.dir=/var/log/sipxpbx -Dorg.apache.lucene.lockdir=/var/sipxdata/tmp/index -Dorg.apache.commons.loging.Log=org.apache.commons.logging.impl.Log4JLogger -Djava.awt.headless=true org.mortbay.jetty.Server /etc/sipxpbx/sipxconfig-jetty.xml
root 6142 0.0 0.0 103240 864 pts/0 S+ 17:47 0:00 grep 5084
[***@sipx ~]#


Note that I have not connected to the machine via HTTP yet - no config has even begun.

I'm going to let this one cook overnight as well...

[Tony Graziano]

I did the same install over the weekend without these kind of issues.

Centos 6.2 minimal install.

Set network info

Yum install epel-release
Yum update
Reboot
Yum groupinstall sipxecs

I did use a vmware environment though. SINCE this is a Linux host, I think
maybe you need a leap second fix for your kernel. Really, I think that's
the problem. Waiting won't help.

Post by Kurt Albershardt
Allocated 3 CPUs and 4GB of RAM to the VM
ran 'yum update' and restarted VM.
Competed 4.6 install per wiki.
Configuring system, this may take a few minutes...
then
for the next ~50 minutes, memory usage on the VM stayed at 2.76GB and CPU
at 52-53% (of 3 CPUs.)
then memory jumped to 3.99GB, java restarted several times with lower CPU and RAM numbers.
done.
5084 sipx 20 0 3014m 546m 13m S 227.0 13.3 9:07.83 java
sipx 5084 211 13.3 3086832 559556 ? Ssl 17:41 11:44
/usr/bin/java -Dcom.ibm.tools.attach.enable=no -Dprocname=sipxconfig
-XX:MaxPermSize=128M -Xmx1024m -Djava.io.tmpdir=/var/sipxdata/tmp
-Djetty.lib.dir=/usr/share/java/sipXecs/sipXconfig
-Djetty.conf.dir=/etc/sipxpbx -Djetty.log.dir=/var/log/sipxpbx
-Dorg.apache.lucene.lockdir=/var/sipxdata/tmp/index
-Dorg.apache.commons.loging.Log=org.apache.commons.logging.impl.Log4JLogger
-Djava.awt.headless=true org.mortbay.jetty.Server
/etc/sipxpbx/sipxconfig-jetty.xml
root 6142 0.0 0.0 103240 864 pts/0 S+ 17:47 0:00 grep 5084
Note that I have not connected to the machine via HTTP yet - no config has even begun.
I'm going to let this one cook overnight as well...
_______________________________________________
sipx-users mailing list
List Archive: http://list.sipfoundry.org/archive/sipx-users/

[Kurt Albershardt]

Post by Tony Graziano
I did the same install over the weekend without these kind of issues.
Centos 6.2 minimal install.
Set network info
Yum install epel-release
Yum update
Reboot
Yum groupinstall sipxecs
I did use a vmware environment though. SINCE this is a Linux host, I think maybe you need a leap second fix for your kernel. Really, I think that's the problem. Waiting won't help.

I didn't realize that a system built after July 1st and running ntp would have leap second vulnerabilities.

I have experience with precision time broadcast television in the early 1980s and then a decade of working with SONET. Guess I don't understand some of the subtleties the *nix kernel brings to the table here.

Could this be some kind of issue between clocks on the VM host and the child VM?

[Michael Picher]

You should always do a yum update to make sure the OS is up to date before
a sipXecs install (if installing from RPM) or directly after install if
installing from ISO.

Also, fyi and just so folks remember, the ISO is not re-created with every
build of the code, only when there is a release that the development team
feels it's necessary. So a 'yum update' is very critical for ISO installs.

Mike

Post by Tony Graziano
I did the same install over the weekend without these kind of issues.
Centos 6.2 minimal install.
Set network info
Yum install epel-release
Yum update
Reboot
Yum groupinstall sipxecs
I did use a vmware environment though. SINCE this is a Linux host, I think
maybe you need a leap second fix for your kernel. Really, I think that's
the problem. Waiting won't help.
I didn't realize that a system built after July 1st and running ntp would
have leap second vulnerabilities.
I have experience with precision time broadcast television in the early
1980s and then a decade of working with SONET. Guess I don't understand
some of the subtleties the *nix kernel brings to the table here.
Could this be some kind of issue between clocks on the VM host and the child VM?
_______________________________________________
sipx-users mailing list
List Archive: http://list.sipfoundry.org/archive/sipx-users/

--
Michael Picher, Director of Technical Services
eZuce, Inc.

300 Brickstone Square****

Suite 201****

Andover, MA. 01810
O.978-296-1005 X2015
M.207-956-0262
@mpicher <http://twitter.com/mpicher>
linkedin <http://www.linkedin.com/profile/view?id=35504760&trk=tab_pro>
www.ezuce.com

------------------------------------------------------------------------------------------------------------
There are 10 kinds of people in the world, those who understand binary and
those who don't.

[Kurt Albershardt]

Post by Kurt Albershardt
Allocated 3 CPUs and 4GB of RAM to the VM
ran 'yum update' and restarted VM.
Completed 4.6 install per wiki.

Still trying to understand how a leap second issue might affect a system installed, updated, and restarted _after_ July 1st?

Post by Kurt Albershardt
You should always do a yum update to make sure the OS is up to date before a sipXecs install (if installing from RPM) or directly after install if installing from ISO.
Also, fyi and just so folks remember, the ISO is not re-created with every build of the code, only when there is a release that the development team feels it's necessary. So a 'yum update' is very critical for ISO installs.
Mike

Post by Tony Graziano
I did the same install over the weekend without these kind of issues.
Centos 6.2 minimal install.
Set network info
Yum install epel-release
Yum update
Reboot
Yum groupinstall sipxecs
I did use a vmware environment though. SINCE this is a Linux host, I think maybe you need a leap second fix for your kernel. Really, I think that's the problem. Waiting won't help.

I didn't realize that a system built after July 1st and running ntp would have leap second vulnerabilities.
I have experience with precision time broadcast television in the early 1980s and then a decade of working with SONET. Guess I don't understand some of the subtleties the *nix kernel brings to the table here.
Could this be some kind of issue between clocks on the VM host and the child VM?
_______________________________________________
sipx-users mailing list
List Archive: http://list.sipfoundry.org/archive/sipx-users/
--
Michael Picher, Director of Technical Services
eZuce, Inc.
300 Brickstone Square
Suite 201
Andover, MA. 01810
O.978-296-1005 X2015
M.207-956-0262
@mpicher <http://twitter.com/mpicher>
linkedin
www.ezuce.com
------------------------------------------------------------------------------------------------------------
There are 10 kinds of people in the world, those who understand binary and those who don't.
_______________________________________________
sipx-users mailing list
List Archive: http://list.sipfoundry.org/archive/sipx-users/

[George Niculae]

Post by Kurt Albershardt
Allocated 3 CPUs and 4GB of RAM to the VM
ran 'yum update' and restarted VM.
Completed 4.6 install per wiki.
Still trying to understand how a leap second issue might affect a system
installed, updated, and restarted _after_ July 1st?

Was the VM created before July 1st? I think run into something similar
when I reverted a snapshot / re installed 4.6.0. Check if ntpd up and
running and time in sync

George

[Kurt Albershardt]

Post by George Niculae

Post by Kurt Albershardt
Still trying to understand how a leap second issue might affect a system
installed, updated, and restarted _after_ July 1st?

Was the VM created before July 1st? I think run into something similar
when I reverted a snapshot / re installed 4.6.0. Check if ntpd up and
running and time in sync

VM created July 10th using an OpenVZ template (definitely created earlier.) System updated and restarted July 10th.

The VM host (Proxmox) was created in April and has not been restarted recently.

[Tony Graziano]

Oh ye of little faith.

http://forum.proxmox.com/threads/10165-problem-with-ntpd-(leap-second-day-bug)

Re: problem with ntpd (leap second day bug)
Originally Posted by hamed
There seems be to a bug with ntpd service because of the leap second day.
http://serverfault.com/questions/403732/anyone-else-experiencing-high-rates-of-linux-server-crashes-during-a-leap-second

One of my nodes is affected by this bug. The above link suggested that
we stop ntpd service for now.
Does stopping ntpd service cause any issues for proxmox?

Thanks
No problems here but reports have stated that RHEL6 have had some
issues regarding the leap second. Those reports also state that the
following command sequence should solve the problem:
date ; sudo date -s "date -u" ; date

Post by Kurt Albershardt

Post by George Niculae

Post by Kurt Albershardt
Still trying to understand how a leap second issue might affect a system
installed, updated, and restarted _after_ July 1st?

Was the VM created before July 1st? I think run into something similar
when I reverted a snapshot / re installed 4.6.0. Check if ntpd up and
running and time in sync

VM created July 10th using an OpenVZ template (definitely created earlier.) System updated and restarted July 10th.
The VM host (Proxmox) was created in April and has not been restarted recently.
_______________________________________________
sipx-users mailing list
List Archive: http://list.sipfoundry.org/archive/sipx-users/

--
~~~~~~~~~~~~~~~~~~
Tony Graziano, Manager
Telephone: [phone]
sip: ***@voice.myitdepartment.net
Fax: [phone]
~~~~~~~~~~~~~~~~~~
Linked-In Profile:
http://www.linkedin.com/pub/tony-graziano/14/4a6/7a4
Ask about our Internet Fax services!
~~~~~~~~~~~~~~~~~~

Using or developing for sipXecs from SIPFoundry? Ask me about sipX-CoLab 2013!

[Kurt Albershardt]

Post by Tony Graziano
Oh ye of little faith.
http://forum.proxmox.com/threads/10165-problem-with-ntpd-(leap-second-day-bug)
Re: problem with ntpd (leap second day bug)
Originally Posted by hamed
There seems be to a bug with ntpd service because of the leap second day.
http://serverfault.com/questions/403732/anyone-else-experiencing-high-rates-of-linux-server-crashes-during-a-leap-second
One of my nodes is affected by this bug. The above link suggested that
we stop ntpd service for now.
Does stopping ntpd service cause any issues for proxmox?
Thanks
No problems here but reports have stated that RHEL6 have had some
issues regarding the leap second. Those reports also state that the
date ; sudo date -s "date -u" ; date

This may indeed be an issue between the VM host and the VM. I got to thinking about NTP architecture in a VM environment last night - wondering why run NTP on each individual VM when the host is providing the 'hardware clock' the VMs use.
you do not need a ntp server in a container (and its not possible) - time comes from the host.

The 'not possible' part concerns me a bit so I'm going to ask for more info.

[Kurt Albershardt]

Post by Kurt Albershardt

Post by Tony Graziano
There seems be to a bug with ntpd service because of the leap second day.
http://serverfault.com/questions/403732/anyone-else-experiencing-high-rates-of-linux-server-crashes-during-a-leap-second

This may indeed be an issue between the VM host and the VM. I got to thinking about NTP architecture in a VM environment last night - wondering why run NTP on each individual VM when the host is providing the 'hardware clock' the VMs use.
you do not need a ntp server in a container (and its not possible) - time comes from the host.
The 'not possible' part concerns me a bit so I'm going to ask for more info.

Rebooting the host cleared the deadlock -- CPU usage is now minimal at rest.

Turns out that OpenVZ (by design) prevents client VMs from changing system time. Their position is to enable this on one container only and not to run NTP on the host
http://download.swsoft.com/virtuozzo/virtuozzo4.0/docs/en/lin/VzLinuxUG/330.htm explains the OpenVZ
BUT
Proxmox runs NTP on the host by default.
SO
Rebooting and allowing NTP to update the VM actually did nothing, since the CT was running on host time, which had not been rebooted.


I can disable NTP on the CT, but if sipxecs (at least in the groupinstall) has a dependency on ntp, I'd prefer to configure around it so I don't get overwritten on a future update. Preferred method?


--thanks

[Douglas Hubler]

Post by Kurt Albershardt

Post by Kurt Albershardt

Post by Tony Graziano
There seems be to a bug with ntpd service because of the leap second day.

http://serverfault.com/questions/403732/anyone-else-experiencing-high-rates-of-linux-server-crashes-during-a-leap-second

Post by Kurt Albershardt

Post by Kurt Albershardt
This may indeed be an issue between the VM host and the VM. I got to

thinking about NTP architecture in a VM environment last night - wondering
why run NTP on each individual VM when the host is providing the 'hardware
clock' the VMs use.

Post by Kurt Albershardt

Post by Kurt Albershardt
you do not need a ntp server in a container (and its not possible) -

time comes from the host.

Post by Kurt Albershardt

Post by Kurt Albershardt
The 'not possible' part concerns me a bit so I'm going to ask for more info.

Rebooting the host cleared the deadlock -- CPU usage is now minimal at rest.
Turns out that OpenVZ (by design) prevents client VMs from changing

system time. Their position is to enable this on one container only and
not to run NTP on the host
http://download.swsoft.com/virtuozzo/virtuozzo4.0/docs/en/lin/VzLinuxUG/330.htmexplains
the OpenVZ

Post by Kurt Albershardt
BUT
Proxmox runs NTP on the host by default.
SO
Rebooting and allowing NTP to update the VM actually did nothing, since

the CT was running on host time, which had not been rebooted.

Post by Kurt Albershardt
I can disable NTP on the CT, but if sipxecs (at least in the

groupinstall) has a dependency on ntp, I'd prefer to configure around it so
I don't get overwritten on a future update. Preferred method

In ntp settings there's an option for unmanaged just for this case

Post by Kurt Albershardt
--thanks
_______________________________________________
sipx-users mailing list
List Archive: http://list.sipfoundry.org/archive/sipx-users/

[Kurt Albershardt]

Post by Douglas Hubler

Post by Kurt Albershardt
I can disable NTP on the CT, but if sipxecs (at least in the groupinstall) has a dependency on ntp, I'd prefer to configure around it so I don't get overwritten on a future update. Preferred method

In ntp settings there's an option for unmanaged just for this case

Choosing the Date/Time menu produced

An internal error has occurred. Click here <https://sipx.domain.com/sipxconfig/restart.svc> to continue.

[Douglas Hubler]

Post by Kurt Albershardt
Choosing the Date/Time menu produced
An internal error has occurred. Click here
<https://sipx.domain.com/sipxconfig/restart.svc> to continue.

I could reproduce this, i created
http://track.sipfoundry.org/browse/XX-10262

I'll take a look

[Douglas Hubler]

Post by Douglas Hubler

Post by Kurt Albershardt
Choosing the Date/Time menu produced
An internal error has occurred. Click here
<https://sipx.domain.com/sipxconfig/restart.svc> to continue.

I could reproduce this, i created
http://track.sipfoundry.org/browse/XX-10262
I'll take a look

I pushed a fix for centos 64 bit, page renders, but page still has
some issues. You should be able to get to unmanaged setting though.

push for other distros will get pushed in next hr. or so.

NOTE: I've starting pushing to yet another directory

http://download.sipfoundry.org/pub/sipXecs/4.6.0

So please update your /etc/yum.repos/sipxecs.repo accordingly. I
updated wiki page.

[Kurt Albershardt]

Post by Douglas Hubler

Post by Douglas Hubler

Post by Kurt Albershardt
Choosing the Date/Time menu produced
An internal error has occurred. Click here
<https://sipx.domain.com/sipxconfig/restart.svc> to continue.

I could reproduce this, i created
http://track.sipfoundry.org/browse/XX-10262
I'll take a look

I pushed a fix for centos 64 bit, page renders, but page still has
some issues. You should be able to get to unmanaged setting though.

Check - worked as expected here.

Thanks~

[Kurt Albershardt]

Post by Kurt Albershardt

Post by Douglas Hubler

Post by Douglas Hubler

Post by Kurt Albershardt
Choosing the Date/Time menu produced
An internal error has occurred. Click here
<https://sipx.domain.com/sipxconfig/restart.svc> to continue.

I could reproduce this, i created
http://track.sipfoundry.org/browse/XX-10262
I'll take a look

I pushed a fix for centos 64 bit, page renders, but page still has
some issues. You should be able to get to unmanaged setting though.

Check - worked as expected here.

Two days ago, but fails again today. Perhaps choosing 'unmanaged service' did something?

[Tony Graziano]

ensure you have the repo pointed to the correct location and do an
update to make sure you have the latest.

Post by Kurt Albershardt

Post by Kurt Albershardt

Post by Douglas Hubler

Post by Douglas Hubler

Post by Kurt Albershardt
Choosing the Date/Time menu produced
An internal error has occurred. Click here
<https://sipx.domain.com/sipxconfig/restart.svc> to continue.

I could reproduce this, i created
http://track.sipfoundry.org/browse/XX-10262
I'll take a look

I pushed a fix for centos 64 bit, page renders, but page still has
some issues. You should be able to get to unmanaged setting though.

Check - worked as expected here.

Two days ago, but fails again today. Perhaps choosing 'unmanaged service' did something?
_______________________________________________
sipx-users mailing list
List Archive: http://list.sipfoundry.org/archive/sipx-users/

--
~~~~~~~~~~~~~~~~~~
Tony Graziano, Manager
Telephone: [phone]
sip: ***@voice.myitdepartment.net
Fax: [phone]
~~~~~~~~~~~~~~~~~~
Linked-In Profile:
http://www.linkedin.com/pub/tony-graziano/14/4a6/7a4
Ask about our Internet Fax services!
~~~~~~~~~~~~~~~~~~

Using or developing for sipXecs from SIPFoundry? Ask me about sipX-CoLab 2013!

[Kurt Albershardt]

Post by Tony Graziano
ensure you have the repo pointed to the correct location and do an
update to make sure you have the latest.

Ran an update yesterday PM and got nothing new - has the repo location changed since

Post by Tony Graziano
NOTE: I've starting pushing to yet another directory
http://download.sipfoundry.org/pub/sipXecs/4.6.0
So please update your /etc/yum.repos/sipxecs.repo accordingly. I
updated wiki page.

?

[Kurt Albershardt]

Post by Tony Graziano
ensure you have the repo pointed to the correct location and do an
update to make sure you have the latest.

Ran an update yesterday PM and got nothing new - has the repo location changed since Jul 10 ?

Just built a new system and did a clean install of 4.6 per the wiki. Running "rpm -qa | grep | sort" and diffing the outputs between the two systems showed them to be the same.

Reconfiguring from scratch, will see what turns up.

[Kurt Albershardt]

Just an FYI -- after installing today's package updates on a 4.6 system which has NTP configured as an unmanaged service, I get:

Running rpm_check_debug
ERROR with rpm_check_debug vs depsolve:
ntp is needed by sipxsupervisor-4.6.0-104.g90aeb.x86_64
** Found 1 pre-existing rpmdb problem(s), 'yum check' output follows:
sipxsupervisor-4.6.0-84.g2b8f3.x86_64 has missing requires of ntp
Your transaction was saved, rerun it with: yum load-transaction /tmp/yum_save_tx-2012-07-17-16-518KU_x6.yumtx
[***@sipx ~]#

[Kurt Albershardt]

On the fresh 4.6 install I just started up, the System / Date and Time menu still produces:

An internal error has occurred. Click here to continue.

The ntp package is installed on the system, but ntpd will not start due to the OpenVZ restrictions I mentioned a few days back.

[Kurt Albershardt]

Still seeing this in the latest builds, FYI.

Post by Kurt Albershardt
An internal error has occurred. Click here to continue.
The ntp package is installed on the system, but ntpd will not start due to the OpenVZ restrictions I mentioned a few days back.

[Kurt Albershardt]

Post by Tony Graziano
ensure you have the repo pointed to the correct location and do an
update to make sure you have the latest.

Ran an update yesterday PM and got nothing new - has the repo location changed since July 10th?

Another yum update today, but no sipx packages were updated.

Am I querying the wrong repo?

[Kurt Albershardt]

Post by Kurt Albershardt

Post by Tony Graziano
ensure you have the repo pointed to the correct location and do an
update to make sure you have the latest.

Ran an update yesterday PM and got nothing new - has the repo location changed since July 10th?

Another yum update today, but no sipx packages were updated.
Am I querying the wrong repo?

Apparently not - just updated a restored VM to test DNS config and saw a bunch of 4.6.0-99 and 4.6.0-100 packages.

thanks~

[Kurt Albershardt]

Post by Kurt Albershardt
Choosing the Date/Time menu produced
An internal error has occurred. Click here <https://sipx.domain.com/sipxconfig/restart.svc> to continue.

Got the same error page when choosing "Basic Network" from the lefthand side of a phone device page. No problem with any other items on the lefthand side of that page. Same behavior on other phone pages.

Running sipxconfig-4.6.0-72.g5c7e4.x86_64 (is this the correct package to cite?)

[Douglas Hubler]

Post by Kurt Albershardt
Got the same error page when choosing "Basic Network" from the lefthand side
of a phone device page. No problem with any other items on the lefthand
side of that page. Same behavior on other phone pages.

I could duplicate and I captured here
http://track.sipfoundry.org/browse/XX-10263

[Jeff Gilmore]

Kurt,
I'm glad you got things working finally!

I'm interested in the results of your testing, as I too use Proxmox for other servers and would love to host my modest 25 user Sipx there as well. If you would share your experiences that would be appreciated.

Jeff Gilmore

Post by Kurt Albershardt

Post by Kurt Albershardt

Post by Tony Graziano
There seems be to a bug with ntpd service because of the leap second day.
http://serverfault.com/questions/403732/anyone-else-experiencing-high-rates-of-linux-server-crashes-during-a-leap-second

This may indeed be an issue between the VM host and the VM. I got to thinking about NTP architecture in a VM environment last night - wondering why run NTP on each individual VM when the host is providing the 'hardware clock' the VMs use.
you do not need a ntp server in a container (and its not possible) - time comes from the host.
The 'not possible' part concerns me a bit so I'm going to ask for more info.

Rebooting the host cleared the deadlock -- CPU usage is now minimal at rest.
Turns out that OpenVZ (by design) prevents client VMs from changing system time. Their position is to enable this on one container only and not to run NTP on the host
http://download.swsoft.com/virtuozzo/virtuozzo4.0/docs/en/lin/VzLinuxUG/330.htm explains the OpenVZ
BUT
Proxmox runs NTP on the host by default.
SO
Rebooting and allowing NTP to update the VM actually did nothing, since the CT was running on host time, which had not been rebooted.
I can disable NTP on the CT, but if sipxecs (at least in the groupinstall) has a dependency on ntp, I'd prefer to configure around it so I don't get overwritten on a future update. Preferred method?
--thanks
_______________________________________________
sipx-users mailing list
List Archive: http://list.sipfoundry.org/archive/sipx-users/

[Kurt Albershardt]

I'm hoping to put together something just like that, which should apply to OpenVZ in general. Eventually, I'd like to see a sipxecs appliance available which I suspect has the potential to drive a LOT of adoption.

The low overhead of OpenVZ (and template virtualization in general) is a good fit with VoIP.


Now prepare for some newbie questions as I get the system actually running...

Maybe a brief wiki page on the do"s and dont's with proxmox and state your user and media server load so people have the proper expectations if they try this also.
Kurt,
I'm glad you got things working finally!
I'm interested in the results of your testing, as I too use Proxmox for other servers and would love to host my modest 25 user Sipx there as well. If you would share your experiences that would be appreciated.
Jeff Gilmore

[Tony Graziano]

Maybe a brief wiki page on the do"s and dont's with proxmox and state your
user and media server load so people have the proper expectations if they
try this also.

Post by Tony Graziano

Post by Kurt Albershardt

Post by Tony Graziano
There seems be to a bug with ntpd service because of the leap second

day.
http://serverfault.com/questions/403732/anyone-else-experiencing-high-rates-of-linux-server-crashes-during-a-leap-second

Post by Kurt Albershardt
This may indeed be an issue between the VM host and the VM. I got to

thinking about NTP architecture in a VM environment last night - wondering
why run NTP on each individual VM when the host is providing the 'hardware
clock' the VMs use.

Post by Kurt Albershardt

Post by Tony Graziano
From http://forum.proxmox.com/threads/986-NTP-and-proxmox (by one of

you do not need a ntp server in a container (and its not possible) -

time comes from the host.

Post by Kurt Albershardt
The 'not possible' part concerns me a bit so I'm going to ask for more

info.
Rebooting the host cleared the deadlock -- CPU usage is now minimal at rest.
Turns out that OpenVZ (by design) prevents client VMs from changing system
time. Their position is to enable this on one container only and not to
run NTP on the host
http://download.swsoft.com/virtuozzo/virtuozzo4.0/docs/en/lin/VzLinuxUG/330.htmexplains the OpenVZ
BUT
Proxmox runs NTP on the host by default.
SO
Rebooting and allowing NTP to update the VM actually did nothing, since
the CT was running on host time, which had not been rebooted.
I can disable NTP on the CT, but if sipxecs (at least in the groupinstall)
has a dependency on ntp, I'd prefer to configure around it so I don't get
overwritten on a future update. Preferred method?
--thanks
_______________________________________________
sipx-users mailing list
List Archive: http://list.sipfoundry.org/archive/sipx-users/

[Kurt Albershardt]

Post by Tony Graziano
SINCE this is a Linux host, I think maybe you need a leap second fix for your kernel. Really, I think that's the problem. Waiting won't help.

There do seem to be a number of reports of Java processes hanging.

https://lwn.net/Articles/504658/ has a good discussion.