Hi all

I have a problem with a windows prcess killing our Interplay/Unity environment.

We have 7 Avid 3.1.2 PC (HP XW8600) workstations all connected to Unity 5 and Interplay 1.2.4 On one of the avid there's a process called ummdworker.exe which is constantly chunking badwidth on the system, sometimes up to 200 MB/s. The system has been re-installed and the problem persists. I tried googling this process and didn't come up with anything usefull.

Any suggestions?

Thanks,

Oren

I believe that process belongs to the media indexer service.

Thanks for the reply.

Do you know what to fix and how?

Thanks

Oren

Hi Oren,

The ummdworker.exe is a thread that runs when the Media Indexer is scanning storage. This is the main purpose of the Media Indexer. You will usually see as many instances as you have processors. Each thread runs per processor. Indexing storage is a resource heavy task. However, that being said, if it is degrading the performance of your network, then it's likely that there is either a threshold that has been crossed, or a misconfiguration somewhere in the infrastructure.

If the process is using up too much throughput, it could mean that the MI or Lookup Services are misconfigured. It's also possible that there is a problem with your network. It is common for throughput issues to occur when the workspaces are incorrectly mounted in MediaNet configurations. (The MI should be mounted via fibre, however if the configuration is done improperly, it is possible for them to try to mount both via fibre and ethernet.)

Is the problem occurring on the client systems or on the Media Indexer server?

You also mention that you have seven clients running the 3.1.2 editing software, on a 1.2.4 back end.

What is the version of the Interplay software that is installed on the client? (Access, Media Indexer Client, Interplay Framework for Client) Per the Backwards Compatibility matrix, the Interplay software that is installed on the client systems must match the release build version for the editing client, not the server version. This means that you should have Interplay 1.5 software on the client systems.

What Framework services are running on the client systems? We should see only Machine Monitor, Service Starter, Time Sync and Media Indexer. (If you are running the 1.5 software you may see Media Indexer and Workstation Service). If you are running the Lookup Service on client machines, I would expect excessive network traffic and perhaps resource contention, but not from the ummdworker.exe threads. (However, if you are running the Lookup Service or System Configuration service on the clients, the configuration is incorrect and needs to be corrected for proper operation. These are master services.)

How many servers are in your configuration? Do you have a dedicated Media Indexer and Interplay Engine? Do You have a combo server running all of the services and applications? Do you have multiple indexers (are any of them delegates for dedicated indexing?)? Is there a dedicated server for the Framework master services?

How many files are on your shared storage? If you are running with a dedicated 32-bit Media Indexer in a HAG, you have a maximum file count of 250,000 files. If you are running a combo server, with less than 10 clients, you can achieve up to 400,000 files.

Are you indexing local storage? If so, is the Media Table entry reflected in the local client Media Indexer tables where the local storage is located (proper config)? or is it located in the Master Media Table from the instance of the MI server (incorrect configuration for local storage.)? Indexing a client's local storage from the MI server can lead to circular references in the ummd tables. This loop condition can cause excessive resource usage and excessive throughput consumption. The consequences would be akin to a routing loop or switching loop/spanning tree problem on a network, where the ummd tables can become corrupt, and even the potental for denial of service.

Have you opened up a case with Avid Customer Success at 1-800-NEWS-DNG? I would highly recommend opening up a case with Customer Success if this is causing infrastructure problems.

Ed

Thanks Ed!

We got the local Avid partner involved and they're on the case.

Thanks for the detailed reply.

Oren