Hi,
In a scenario in which a DMA tries to communicate with a remote DMP and the network latency is suboptimal (>500 ms), is there any mechanism we can use to allow SLNet to be more “tolerant” and wait a bit longer for a response to a request?
Scenario:
The following error is issued when trying to upload an upgrade package:
Notices
– No longer checking for progress from http://<DMP IP>:8004/UpgradeService: too long since last event received
When you have a DMP where the network is suboptimal i would recommend to only communicate with replication, which contains a buffer when the network is no longer available or has a short timeout
Related to upgrades, i would suggest to upload first the package to the DMP itself and launch the upgrade from the DMP instead of using the main cluster for launching the upgrade