[
https://issues.apache.org/jira/browse/CASSANDRA-11383?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15202206#comment-15202206
]
DOAN DuyHai commented on CASSANDRA-11383:
-----------------------------------------
[~jkrupan]
1. --> Yes correct
2. Originally we targeted 4 billions but we stop after 80 iterations (instead
of 100) so it gave us something like 3.4 billions
3. Yes correct. But this territory_code is not meant to be used alone but in
combination with the resource_period_end_month_int_idx index and others to cut
down the number of rows to be fetched
4. and 5. Same answer as above
Indeed, those indices are designed to support an user-search form to perform
TopK aggregation with dynamic filtering.
SASI is used for the filtering part and Spark for the TopK aggregation
Indeed, the *resource_period_end_month* filter is mandatory for the user and
usually people put a month or a range of 1 year at most. This cuts down the
whole dataset to a reasonable subset over which we apply other filters
> SASI index build leads to massive OOM
> -------------------------------------
>
> Key: CASSANDRA-11383
> URL: https://issues.apache.org/jira/browse/CASSANDRA-11383
> Project: Cassandra
> Issue Type: Bug
> Components: CQL
> Environment: C* 3.4
> Reporter: DOAN DuyHai
> Attachments: CASSANDRA-11383.patch, new_system_log_CMS_8GB_OOM.log,
> system.log_sasi_build_oom
>
>
> 13 bare metal machines
> - 6 cores CPU (12 HT)
> - 64Gb RAM
> - 4 SSD in RAID0
> JVM settings:
> - G1 GC
> - Xms32G, Xmx32G
> Data set:
> - ≈ 100Gb/per node
> - 1.3 Tb cluster-wide
> - ≈ 20Gb for all SASI indices
> C* settings:
> - concurrent_compactors: 1
> - compaction_throughput_mb_per_sec: 256
> - memtable_heap_space_in_mb: 2048
> - memtable_offheap_space_in_mb: 2048
> I created 9 SASI indices
> - 8 indices with text field, NonTokenizingAnalyser, PREFIX mode,
> case-insensitive
> - 1 index with numeric field, SPARSE mode
> After a while, the nodes just gone OOM.
> I attach log files. You can see a lot of GC happening while index segments
> are flush to disk. At some point the node OOM ...
> /cc [~xedin]
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
