Hello, I deployed signoz via helm chart in k8s clu...
# support
l
Hello, I deployed signoz via helm chart in k8s cluster, and I checked its clickhouse table schema, found that it's based on the
MergeTree
engine, not the
ReplicatedMergeTree
, is it by design or have I misconfigured ?
s
If you need replication you need to explicitly enable it by setting
Copy code
schemaMigrator:
  enableReplication: true
l
Thanks, I'll try
Hi @Srikanth Chekuri I enabled the replication and rerun the migrator job, but the engine still is MergeTree
do i need to recreate the clickhouse cluster?
s
Yes
l
Hi I just met another issue of the trace data, seems the data of
signoz_index_v2
falls behind that of
signoz_spans
Note the timestamp column of the two tables:
even the data of
signoz_spans
is lagged, the query happens at about 032800 but the newest data is 031326
s
Your question is not clear. It can't lag behind several minutes. It means either data is not there or got dropped.
l
I re-deployed the clickhouse cluster, now it works normally now.
Can I control the batch size or batch timeout of the clickhouse trace exporter? I'd like to configure large batch size to increase the insert performance
I find the config for batch processor, thanks