[Test environment]
- AWS EC2 Instance : c4.xlarge (cpu 4core, memory 7.5GB)
- Storage(per node) : Data (100GB, 300IOPS), Index (150GB, 426IOPS) / General Purpose SSD (GP2)
- 3 Nodes / replica 1
- couchbase community 4.0.0
[Test]
*create indexes (Global Secondary Index) : 8 EA (include PK)
-
insert 100 million items (item size : 1~4KB)
* Issue#1 When insert 10milion items, server load severe and hangup…. ("%wa" is 70% ~ 90%) * Issue#2 Node's status is often changed to "Pend" * Issue#3 Index warning message appear often. (And indexer proccess is not running "9105, 9102".) "Warning: We are having troubles communicating to the indexer process. The information might be sta.te." * Issue#4 When performing a query by using GSI, there is no result. (But…..Obviously, there are data.)
How do I order to resolve the above issues?
Does the system need to scale up or out?
Please help me…