Windows 2022 DNS Server High Memory Utilization

Question

Windows 2022 DNS Server High Memory Utilization

Answer 1

Hello

Thank you for posting in Microsoft Community forum.

I understand that you are experiencing high memory utilization on your Windows Server 2022 DNS server.

There are several reasons why this could be happening, including:

1. DNS cache size is too large
2. DNS server is under heavy load
3. Memory leak in the DNS server process

To troubleshoot this issue, I recommend the following steps:

1. Check the DNS cache size and reduce it if necessary. You can do this by opening the DNS Manager, right-clicking on the server name, and selecting Properties. Then, go to the Advanced tab and adjust the cache size as needed.
2. Check the DNS server's performance counters to see if it is under heavy load. You can do this by opening the Performance Monitor, adding the DNS Server object, and monitoring the various counters.
3. Check for any memory leaks in the DNS server process. You can do this by using tools like Process Explorer or Task Manager to monitor the memory usage of the DNS server process over time.

Best Regards,

Wesley Li

Answer 2

Followed your steps but did not see cache size

Answer 3

I do apologize, but that is not the correct location of what you should be looking for. It's in a separate

location and is not on the properties pages of DNS server.

In case you're still looking:

1. Login to the desktop whether via RDP or console
2. Click on Start > type "performance" and hit . You should see the mmc console pop up.
3. Drill down: Performance > Monitoring Tools > Performance Monitor > +"Properties"

4. Click on the *Add* button and don't just go for what's being request (above), really add some metrics to the performance monitor like UDP datagram packet errors, drops, send/receive errors, disk i/o, CPU temperature, bandwidth measurements of the network port attached to the server. There are metrics available for virtual memory, cache memory, page faults, etc.

Note: If you look at the metrics I selected, I toggled the ones around the failure metrics so I could identify them easier. There are a bunch of IPv6 queries failing, but that's due to me removing the dual-stack setup on this LAN. One big thing that I sometimes forget is if I'm hammering the system, the massive I/O will cause the slowdowns for ingress and egress.

I wasn't sure if you already have investigated disk i/o? (there was no mention of network bandwidth previously either)

I'm happy to work with you, and I'll be monitoring this thread for your reply. Take care!

Regards,

Gabriel