Horde Perforce Issues and Metrics Questions

Development Pipeline & Plugins
, ,
1

Telemetry and Metric Views :

- Is it possible to display the UserID/UserSession within aggregated metric views?

         \- Use case : when viewing the PIE boot time metric on our Editor Telemetry page we noticed a single outlier data point. Being able to identify which user or telemetry session this data point originates from, will allow us to support this user directly.

- How to view the metric data to build our own telemetry filters?

         \- Attempts were made to view the raw telemetry data to create our own filters. I found stores of binarized data while traversing the MongoDB, is there a suggested method for identifying what keys and values are present within the telemetry data stores which will allow us to build our own filters? (Specifically, the *Payload.{EntryName}* entries.)

Perforce usage and workspace conform:

- Is it possible to reduce the frequency of idle agent Perforce workspace conform?

         \- It appears agents will periodically perform a Perforce intensive workspace conform. Is it possible to disable this behavior when the agents are idle and specify workspace conform only conducted during job start if desired?

- We encountered an issue when a workspace file, which was not in an expected file format, triggered constant p4 queries from each agent, blocking our servers until the issue was identified.

         \- Details within Reprod Steps.
2

Steps to Reproduce

After setting up Horde with x stream.

Submit unicode file to Perforce unicode files (specifically files that p4 identified as unicode).

Notice Perforce Server being constantly queried by Horde Agents.

3

Hey there,

Apologies for the delay in response here.

> - Is it possible to display the UserID/UserSession within aggregated metric views?

At this time, we do not have this type of data. The primary reason is because it is indeed aggregated - so only the metric contents are considered. That being said, you *could* attempt to enable the Telemetry sink (horde server config - snippet below), and from the telemetry call site (here) make sure you have the appropriate user context.

"plugins": { "Analytics":{ "Sinks": { "Mongo": { "Enabled": true, "RetainDays": 1 } } } },I would very much be careful of the load you place on your MongoDB however, as a lot of telemetry is indeed generated. As the Horde Analytics system is quite experimental still, we are still thinking of ways to better manage the data to give individuals more control over where the Horde Analytics and telemetry data live - to empower use cases like this.

Regarding conform frequency - you do have some controls over this via ConformInterval in the pool config - so this could be worth checking.

Regarding your other questions, let me follow up with the team internally to see if this is an issue that’s being tracked.

Kind regards,

Julian