2.6 KiB
Experiencing Performance Issues in Peak Hours
| ID | Creation Date | Assignee | Status |
|---|---|---|---|
| OMPUB-975 | 2023-07-28T16:02:40.000+0800 | 陆秋文 | 已解决 |
Hello,
Your system has been recently deployed in our setup and the ISPs are complaining of packet drops in peak hours. I am attaching the TSG graphs snapshot from the last 48 hours.
Moreover, we also frequently receive the “container firewall restarted” and “Packet TX Drops by applicaton poor performance” alarms on different TSG nodes. The alerts from the last 48 hours have also been attahched.
Please investigate and help us resolve the matter asap.
Regards,
Naeemluqiuwen commented on 2023-07-28T16:09:01.395+0800:
- Query the CPU usages, please use the following Prometheus query in Network Zodiac (Main->APM->Explore), for 48 hours, and provide the snapshot for us. {code:java} rate(container_cpu_user_seconds_total{container="firewall", datacenter="MSH"}[5m]) rate(container_cpu_user_seconds_total{container="firewall", datacenter="PCAP"}[5m]) rate(container_cpu_user_seconds_total{container="firewall", datacenter="TWA"}[5m]) {code}
- Try to enable the overload bypass in TSG-OS, by following the steps: a. login TSG-OS by ssh and run the "tsg-os-cli" to start the configuration tools. b. enable the overload bypass and commit the changes. {code:java} $ sudo tsg-os-cli root@dell-emc-r7525> set template name tsg_traffic_engine_default overload_protection enable yes root@dell-emc-r7525> commit {code} c. wait for the traffic engine restart (need ~120s), quit the "tsg-os-cli", go back to the bash command line and run the diagnosing procedure {code:java} sudo tsg-diagnose-oneshot {code}
Attachments
Attachment: alert-message-2023-07-28+11-35-21.xlsx
alert-message-2023-07-28+11-35-21.xlsx
Attachment: alert-message-2023-07-28+11-36-20.xlsx
alert-message-2023-07-28+11-36-20.xlsx
Attachment: Results_from_Zodiac_280723.rar
Results_from_Zodiac_280723.rar
Attachment: TSG-X-Status-Customed+(Device+Groups).html