We've been running Cato SDP Client version 5.5.0.29237 across our developer fleet for several months with stable latency to our peered cloud VPCs. Post-update to 5.6.0.30145, we're observing significant, intermittent latency spikes in specific regions.
Our typical baseline from our Frankfurt office to the AWS eu-central-1 PoP is 18-22ms RTT. Since the update, we're seeing jumps to 200-450ms lasting for 30-90 seconds, occurring multiple times per hour. Other regions (e.g., US-East) remain unaffected. This is critical for our real-time sync services.
Troubleshooting steps taken:
* Verified local network congestion is not a factor.
* Rolled back to 5.5.0 on a test machine—latency returned to baseline.
* Compared `traceroute` outputs during a spike event. The path appears unchanged, but hop 3 (inside Cato's backbone) shows the dramatic increase.
Configuration is standard:
```json
{
"TunnelMode": "SplitTunnel",
"CloudApplications": ["AWS-us-east-1", "AWS-eu-central-1"],
"MTU": 1400
}
```
Has anyone else performed empirical measurements before/after this update? I'm particularly interested in:
* Whether this is isolated to certain PoP pairings.
* If Cato's support has acknowledged any known issues with their latest transport layer adjustments.
* Workarounds beyond holding back client updates.
benchmark or bust
benchmark or bust