Details
-
Bug
-
Resolution: Duplicate
-
Major
-
3.0
-
Security Level: Public
-
3.0.0-798
Platform = Physical
OS = CentOS 6.5
CPU = Intel Xeon E5-2680 v2 (40 vCPU)
Memory = 256 GB
Disk = RAID 10 SSD
Description
0. 10 nodes
1. Create 20 buckets
2. Wait for cluster to become healthy
3. Sleep 5 minutes
4. Remove all buckets
5. Sleep 2.5 minutes
6. Create 30 buckets
At this point all servers become "pending" and attempt to auto-failover all nodes reported:
Could not auto-failover node ('ns_1@172.23.100.17'). There was at least another node down.
Could not auto-failover node ('ns_1@172.23.100.18'). There was at least another node down.
Could not auto-failover node ('ns_1@172.23.100.19'). There was at least another node down.
Could not auto-failover node ('ns_1@172.23.100.20'). There was at least another node down.
Could not auto-failover node ('ns_1@172.23.100.21'). There was at least another node down.
Could not auto-failover node ('ns_1@172.23.100.22'). There was at least another node down.
Could not auto-failover node ('ns_1@172.23.100.23'). There was at least another node down.
Could not auto-failover node ('ns_1@172.23.100.24'). There was at least another node down.
Could not auto-failover node ('ns_1@172.23.100.25'). There was at least another node down.
Could not auto-failover node ('ns_1@172.23.100.26'). There was at least another node down.
memcached reports "Too many connections" and I have trouble finding anything in 600M+ diag logs.
CBEE 2.5.1 works fine.
Attachments
Issue Links
- blocks
-
MB-10914 {UPR} ::Control connection to memcached on 'ns_1@IP' disconnected with some other crashes before upr_replicator:init/1, upr_proxy:init/1, replication_manager:init/1
- Closed