Running scale test of connected 10K instances, we end up consuming 1.2GB heap, which boils down to around ~125kB per session overhead.
Drilling down into this, it looks like we are retaining a 32KiB HeapByteBuffer in each session – i.e. 25%, through a read timeout task. Investigate the root cause for this, potentially switching to netty provider instead of nio2.
- is blocked by
-
NETCONF-590 Refactor NETCONF transport layer
- Resolved
-
NETCONF-674 Eliminate use of AsyncSshHandlerReader
- Resolved
-
NETCONF-699 Eliminate use of AsyncSshHandlerReader in call home
- Resolved
-
NETCONF-1069 Refactor AbstractNetconfTopology interface
- Resolved
-
NETCONF-1070 Rework client reconnection logic
- Resolved