This post is more than 5 years old
35 Posts
0
3759
February 24th, 2011 03:00
how to suppress the "NTP Servers are not accessible" errors?
My DataMovers have not connectivty to external network. They are connected to two switches dedicated for NFS traffic for vSphere.
So there is no default gateway/route I can set that will allow the DataMover to connect out to a NTP server, as it physically has no connectivty to the rest of our network.
Is there anyway that I can suppress the "Critical" error message about NTP servers not being accessible when I log on to the Control Station?
Thanks.
No Events found!
dynamox
9 Legend
•
20.4K Posts
0
February 26th, 2011 07:00
control station and datamovers are connected via private network. On Control Station run /sbin/ifconfig ..to see the list of internal ip address. Also on control station run cat /etc/hosts. There you will see the ip address on private network assigned to datamovers.
mtg82814
35 Posts
0
February 24th, 2011 04:00
Hi, thanks for response.
The issue I have is that the DataMovers devices (cge0-3) are connected to switches that are dedicated for NFS traffic, those switches are not connected to the rest of my network. So the only thing that connects to them are servers (ESXi hosts) that need to connect to a NFS export. So there are no physical paths for the datamovers to connect to my NTP servers in my network.
That's why I was looking for a way to suppress the error message.
I do have one port free on each of my NFS switches. I suppose I could trunk the switches to my core and turn on dot1q and other fancy network stuff, then I could create a path to my NTP servers, but I would rather not have to go updating all my network infrastructure and then update all the NFS clients (tag interface to be on a NFS vlan) that connect to them if there is a way around it.
Thanks.
rworkman1
26 Posts
0
February 24th, 2011 04:00
I don't know about turning that off on Celerra.....but....
Do you have any servers that can access an external NTP source?
Why not set that one up, and then point your Celerra to that internal server, and have it become your NTP time source.
Just a thought,
Rita
urle
48 Posts
0
February 24th, 2011 05:00
What about deleteing the ntp-entries:
server_date server_2 timesvc stop ntp
server_date server_2 timesvc delete ntp
server_date server_2 timesvc
server_2 : no entries
This will remove the error messages but the times on your movers will disperse over time. Another solution could be to use the controlstation as ntp-source.
Rainer_EMC
4 Operator
•
8.6K Posts
1
February 24th, 2011 12:00
If your control station NTP daemon is working fine you could also configure the data movers to use that over the internal network
Rainer
mtg82814
35 Posts
0
February 26th, 2011 04:00
Hi Rainer, your suggestion sounds like a good idea. Is their a procedure for using the Control station NTP daemon? Do I just set the NTP servers to the IP of the control station. Also if I recall, the Celerra DMs like to see two Datamovers or they will issue an warning, is that right?
Thanks.
mtg82814
35 Posts
0
February 26th, 2011 07:00
Rainer,
I'm not sure what to use for a IP address of the NTP server if I use the Control Station as my NTP server. I obvioulsy can't use the control station IP address as the internal networking has no routes to that IP. Is there an IP address that the internal networking uses to communicate with the control station?
Thanks.
mtg82814
35 Posts
0
February 27th, 2011 15:00
dynamox, that worked. Thanks!
Rainer_EMC
4 Operator
•
8.6K Posts
0
February 28th, 2011 07:00
yes - to find the correct IP adresses look what subnets the mge0/mge1 internal interfaces of the data mover are on and find the corresponding control station internal IP adresses through ifconfig or in /etc/hosts