The % PDC Time (Processor\% DPC Time\_Total) for the system has exceeded the threshold. Overall system performance may significantly diminish which will result in poor operating system and application performance.
% DPC Time is the percentage of time that the processor spends receiving and servicing deferred procedure calls (DPCs). DPCs are interrupts that are run at a lower priority than standard interrupts. If a high % DPC Time is sustained there may be a processor bottleneck or an application or hardware related issue that can significantly diminish overall system performance.
A high % DPC Time value can be caused by one or more of the following:
Hardware or device driver related problem
To determine the root cause of a high DPC, rate follow the process outlined below.
Observe the proportion of the processor time that is spent servicing interrupts and DPCs. To do this, monitor the following counters in real time using System Monitor:
Processor\% Processor Time
Processor\% Interrupt Time
Processor\% DPC Time
If a processor instance is running a sustained % Processor Time that is > 85% and it is also spending > 15% of that time servicing Interrupts and/or DPCs, the processor is probably the source of a performance bottleneck. This bottleneck can be addressed by upgrading or adding additional processors to the computer.
If the processor is running a sustained % Processor Time of < 85% and it is also spending > 15% of that time servicing interrupts and/or DPCs, the performance issue may be the result of either an application or hardware related issue.
Where a hardware device is the root cause, an administrator will find that the % DPC Time has probably increased substantially over a short period time. This will often occur when new hardware is installed or drivers have been upgraded. If the administrator can isolate the issue to a hardware/device driver issue, it can be addressed by working with the vendor.
In cases where you are administering a multiprocessor system that does not distribute interrupts symmetrically, you can often improve the distribution of the processor workload by adding network adapters so that there is one adapter for every processor. Generally, you only add adapters when you need to improve the throughput of your system. Network adapters, like any additional hardware, have some intrinsic overhead. However, if one of the processors is nearly always active (that is, if Processor: % Processor Time = 100) and more than half of its time is spent servicing DPCs (if Processor: % DPC Time > 50), then adding an adapter is likely to improve system performance, as long as the available network bandwidth is not already saturated.
Where an application is the root cause, you will find that DPCs are probably being blocked by an application that has issued a call that is taking a substantial amount of time to complete. During this time DPCs are blocked and will be queued. To determine which application is the root cause, you must run advanced performance, tracing, and diagnostics to pin point the exact application that is responsible for the performance issue.