Summary Of Common Issues And Quick Troubleshooting Methods For Singapore Servers Via Tencent Cloud

1. Quick troubleshooting of common connection and network faults

(1) Check whether the cloud instance's public IP or elastic public IP is bound, and confirm that the security group has opened the corresponding port (such as 80, 443, 22).
(2) Use ping or traceroute to track network paths. Example command: ping -c 6 203.119.XX.XX; traceroute -n 203.119.XX.XX (observe packet loss and hop count).
(3) Within the instance, use netstat -tulpn or ss -tunlp to confirm the service port listening status and verify whether the process corresponds to the port.
(4) If cross-border access is slow, confirm outbound bandwidth and peak usage. It is recommended to check the Cloud Monitor port traffic curve and compare bandwidth quotas.
(5) Troubleshooting routes or ACLs: Check security groups, NAT gateways, subnet routing tables, and network alarm records on the cloud vendor side.

2. CPU, memory, and disk performance issues and solutions

(1) Use top-b-n1 or vmstat to observe instantaneous CPU and load conditions to determine whether usage is a short-term surge or sustained high usage.
(2) Use free -m or cat /proc/meminfo to check memory usage and swap usage; if necessary, expand memory or enable swap.
(3) Disk I/O can be checked using iostat -x 1 3. iotop. If IOPS are saturated, consider switching to CLOUD_SSD or upgrading disk specifications.
(4) Use df -h to confirm the mount point space for disk space, clear logs, or expand the data disk. It is recommended to rotate logs on /var/log.
(5) Example of instance type: CVM S3. MEDIUM (2 vCPU / 4GB memory) is suitable for lightweight applications; CVM C6. HIGH (4 vCPU / 8GB memory) is suitable for small to medium-sized websites.

3. Domain name resolution and DNS troubleshooting

(1) First, use dig or nslookup to check the domain name resolution results and TTL:dig + short example.com @8.8.8.8.
(2) Confirm whether the resolution points to the correct Elastic Public IP (EIP) or Load Balancing Address (CLB), and check for any old cache records.
(3) If using Tencent Cloud DNSPod or hosted domains, check the record type (A/AAAA/CNAME) and the weight settings of the resolution record.
(4) When cross-region access exceptions, check whether geo-routing is enabled or if intelligent parsing rules cause parsing bias toward certain POPs.
(5) For certificate and HTTPS issues, confirm the CN/SAN match between the domain name and certificate, and use openssl s_client -connect domain:443 to view handshake details.

4. CDN acceleration and caching strategies, cache refresh troubleshooting

(1) Confirm that the CDN is enabled and enabled by the domain name, check whether the return to the source is normal, and record the return status code.
(2) Check cache policies (Cache-Control, Expires) and whether version numbers are managed for static resources to avoid cache confusion.
(3) When encountering 5xx or 4xx error in return to the source, review the CDN return log and troubleshoot the status of the backend instance by time.
(4) Example optimization: Set images, JS, and CSS to max-age=31536000 and use a versioned path with a hash filename.
(5) Provides fast refresh commands and console operation paths; in case of emergencies, use the CDN console's refresh interface or API to clean specific URLs.

5. DDoS defense and handling traffic anomalies

(1) Tencent Cloud provides both basic and professional Anti-DDoS products. First, confirm whether the corresponding protection level is enabled and bind the instance or EIP.
(2) When traffic is abnormal, cloud monitoring is used to view the inbound traffic spike curve and record peak bandwidth (for example, a peak of 3200 Mbps).
(3) Identify attack types: SYN flooding, UDP amplification, HTTP flooding, using blacklist and whitelist, request frequency limiting, and WAF strategies respectively.
(4) Real case: During Double 11, an e-commerce company experienced a 1.2 Gbps HTTP flood. After enabling Anti-DDoS Pro, the recovery traffic stabilized at 60 Mbps after cleaning.
(5) Common emergency measures: temporarily increase protection bandwidth, enable CDN cache and WAF, block abnormal source IP segments, and contact cloud vendors for cleaning.

6. Examples of backup, snapshots, monitoring alerts, and real-world configuration

(1) It is recommended to regularly use snapshots or cloud drives for backup, following a 7+7 backup strategy (7-day rolling backups + monthly snapshots).
(2) Cloud Monitor alerts enabled: CPU > 80% for 10 minutes, disk utilization >75%, network incoming traffic > bandwidth threshold.
(3) Example server configuration table (common configurations in Singapore region):

instance name	specification	CPU	memory	bandwidth
web-prod-01	C6. LARGE	4 vCPU	8 GB	100 Mbps
db-prod-01	CVM High IO	8 vCPU	32 GB	dedicated intranet

(4) Real Case Summary: A SaaS customer in Singapore experienced unlimited log disk writes causing disk fullness. Using df -h to confirm /var/log occupying 95% of the data, cleaning old logs and dumping them to object storage, reduced I/O latency by 70%.
(5) Recommended daily SOPs: establish monitoring alerts, periodic snapshots, minimum three-layer protection (security group + WAF + Anti-DDoS), traffic peak drills, and emergency contact procedures.