Performance problems? 2k dynamic dns issue

  • Thread starter Thread starter Kevin W. Gagel
  • Start date Start date
K

Kevin W. Gagel

I have our dns setup with dynamic dns enabled so each dhcp client gets a
record created on startup. The problem I'm seeing is my primary dns will
frequently return a not found error and then if you simply redo the query it
returns the correct results. You can redo the same error within 1 second and
what it could not find it can suddenly return accuratly.

I have webtrends monitoring the dns servers to see if they go down. Every 2
to 15 minutes the server becomes unreachable and then re-appears to the
monitoring software. We've had other software monitoring and it to found
that the primary dns would vanish and come back.

Any pointers on where to look for this?
 
All clients & servers pointing ONLY at the DNS server(s)' LAN IP(s) for DNS
in their IP settings? With *no* external/ISP DNS server IPs specified there?
 
All clients are automatically configured through dhcp which sets them to the
primary and secondary dns servers we have internally. These two servers are
configured with the primary as the one to forward changes to our ISP's dns
servers and our secondary acts as a slave to the primary.

So all clients using dhcp will/do have the primary and secondary dns servers
configured as their dns servers. None of them have the external (ISP)
servers as fall backs.

We have only about 1,000 to 1,500 computers with 400 to 800 active at any
one time. The majority of these are not doing dns queries at the same time.

Simultanious queries could be expected from our labs which operate from 7 to
11 each day for 600 to 700 computers.

The monitoring software is showing the dns servers as unavailable throughout
the night. At this time there will be a maximum of 1 person using the
internet (security). Hence, I'm puzzled about why they don't respond with
the record on the first query but do with the second query.

"Lanwench [MVP - Exchange]"
 
In
Kevin W. Gagel said:
All clients are automatically configured through dhcp which sets them
to the primary and secondary dns servers we have internally. These
two servers are configured with the primary as the one to forward
changes to our ISP's dns servers and our secondary acts as a slave to
the primary.
So all clients using dhcp will/do have the primary and secondary dns
servers configured as their dns servers. None of them have the
external (ISP) servers as fall backs.

We have only about 1,000 to 1,500 computers with 400 to 800 active at
any one time. The majority of these are not doing dns queries at the
same time.
Simultanious queries could be expected from our labs which operate
from 7 to 11 each day for 600 to 700 computers.

The monitoring software is showing the dns servers as unavailable
throughout the night. At this time there will be a maximum of 1
person using the internet (security). Hence, I'm puzzled about why
they don't respond with the record on the first query but do with the
second query.
Click to expand...


Are there any errors in the Event logs of the DNS servers, or any servers at
all?
Are these machines responding to queries from the Internet or are they just
used for your internal AD structure?

Try using Quest's Spotlight on Windows. Download the 30 day trial and
install it on a machine other than the DNS servers and watch the DNS server
performance thru it. You can log the stuff and find out at what time they
drop offline and coorelate that to what in your Event logs or anything else
happening on the network.

--
Regards,
Ace

Please direct all replies ONLY to the Microsoft public newsgroups
so all can benefit.

This posting is provided "AS-IS" with no warranties or guarantees
and confers no rights.

Ace Fekay, MCSE 2003 & 2000, MCSA 2003 & 2000, MCSE+I, MCT, MVP
Microsoft Windows MVP - Windows Server - Directory Services

Security Is Like An Onion, It Has Layers
HAM AND EGGS: A day's work for a chicken;
A lifetime commitment for a pig.
 
Ace,
Thanks for the Spotlight on windows tip. I've downloaded it and run it
against a few of my servers. The DNS server in question shows only that
threads are waiting for the cpu to take care of them. So we'll be upgrading
the box very shortly. Its a strange result because that machine is not under
heavy use when those alarms get sounded. Frequently only 2% or 3% cpu usage
yet the work queue length can be as much as 20 threads waiting on the cpu.


"Ace Fekay [MVP]"
 
Ace,
Thanks for the Spotlight on windows tip. I've downloaded it and run it
against a few of my servers. The DNS server in question shows only
that threads are waiting for the cpu to take care of them. So we'll
be upgrading the box very shortly. Its a strange result because that
machine is not under heavy use when those alarms get sounded.
Frequently only 2% or 3% cpu usage yet the work queue length can be
as much as 20 threads waiting on the cpu.
Click to expand...

Is this on all DNS servers or just the one? If just the one, and the CPU
queue length is greater than 2, it could be pointing to a CPU problem, or an
onboard L2 cache issue on the CPU.

Ace
 
Back
Top