Time Synchronization Error after 60 users login

[Zia]

[img][img][img][img][img]Hi,

I have a set of cluster server, Server A and Server B, A server is installed asterisk and B server is installed DB+Web+Archive services. Vicidial 4.0.3
and VERSION - 2.10-451a, BUILD: 140902-0816. in asterisk server i have installed two 4 port allo PRI card. Both the servers are Dell config as below
Intel(R) Xeon(TM) CPU 3.00GHz
2 CPU each has 4 cores.
RAM 32 GB.

upto 50 to 55 agents server works normally. once users increase 60 to 70 users start getting Error.
" There is Time Synchronization problem in your system, Please tell your System administrator"
and if the agent click on back button normal screen comes but this error is keep repeating every 10 sec.

and i have another set of cluster where 70 to 80 users login and that cluster servers also i3 and 16 GB ram only but its works fine.

Need your help.

Reg:
Zia

[Zia]

When i see the server load not even 30%

[mflorell]

Where are you seeing the server load?

[Zia]

Using TOP utility i am seeing the server ave load.

[mflorell]

is that 30% "30.00" or "0.30" or is it "1.20" on a quad-core server?

Or are you looking at the CPU non-idle%?

[Zia]

In database server its avg load is 3.0 and in asterisk server 0.40 where could be problem

[Zia]

i removed old asterisk server from the cluster and added fresh installed i7 with 12 GB RAM and checked found same issue. once increase agents more than 60 Time Synchronization and same time seen the Top status of server avg load, more or less same load am getting in both the server. i think it might be the problem with database server OR some settings we need to change. Please do help........ this became dilemma for me.

[mflorell]

It is probably a database issue, what are the hardware specs on the DB server(including type of drives and RAID subsystem)?

[Zia]

Intel(R) Xeon(TM) CPU 3.00GHz
2 CPU each has 4 cores.
RAM 32 GB.
1 TB HDD scsi hard disk.

[Zia]

Database repair will it work?

Last Time i faced same error, and found the asterisk server's Hard disk was full. after that i made available some space and repaired the database then started working fine.. but this case is different so much space is there in both the servers. so i am getting crazy,,

[mflorell]

No RAID card?

You say this is a SCSI hard drive, what controller are you using?

[Zia]

Raid controler only ... but we are nt using any raid... only one hard disk... please help me out... it screwed my happiness.

[Zia]

Anyone please help me...

[mflorell]

Have you tried running 'mtop' when you are having problems on your database?

Or turning on slow query logging in mysql/mariadb?

My guess is that your single-drive system is causing the problem. You really need to be using fast drives on a proper RAID card for a database sized such as yours.

[williamconley]

Anyone please help me...

Two likely paths:

1) Time is not synchronized, and heavy load pushes it too far. Are you using NTP? If so, how? What are your settings? Response from all server to this command:

Code: Select all

ntpq -p

2) Heavy load on your local network could be causing dropped packets on your local network routers and/or switches or between the agents and the Vicidial Web server. Thus the problem is not actually Vicidial itself, but dropping network packets will cause that same error.