Splunk Search

Subsearch

yifatcy
Path Finder

Hi,

I've been trying for hours and nothing works, so I figure you might help me out.

I have the following very long query:

 

 

 

| tstats SUM(requested_cpus) as requested_cpus, SUM(reserved_ram) as reserved_ram, SUM(requested_ram) as requested_ram, SUM(used_ram) as used_ram, SUM(compute_ram_total) as compute_ram_total, count as agg_field_seen WHERE (index=monitor (host="$queuename$") fs_group=$fsgroup$ project=$project$ site=$site$            (slave_resource{} IN ("*")) (NOT slave_resource{} IN ("___VALUE_NONE___"))     ( host_state="normal" OR host_state="full" OR host_state="ovrld" OR host_state="sick" OR host_state="susp" OR  host_state="base" OR host_state="ready")     )
          OR (index=ncstat_monitor (host="$queuename$") fs_group=$fsgroup$ project=$project$ site=$site$ (compute_slave_res{} IN ("*")) (NOT compute_slave_res{} IN ("___VALUE_NONE___"))     ( host_state="normal" OR host_state="full" OR host_state="ovrld" OR host_state="sick" OR host_state="susp" OR  host_state="base" OR host_state="ready") ) BY _time, site, fs_group span=15min | eval query_enabled=1
                | eventstats sum(agg_field_seen) AS sum_agg_field_seen BY fs_group
                | sort 0 - sum_agg_field_seen
                | streamstats dc(fs_group) AS rank
                | eval agg_field_ranked=if(rank <= 50 - 1, 'fs_group', "Other")
                | rename agg_field_ranked as fs_group
          | stats SUM(reserved_ram) as reserved_ram, SUM(requested_ram) as requested_ram, SUM(used_ram) as used_ram, SUM(compute_ram_total) as compute_ram_total, SUM(requested_cpus) as requested_cpus BY _time, site, fs_group
| eval slots=max(reserved_ram/32,requested_cpus) | eval full_fsgroup=site.":".fs_group 
          | timechart span=15min limit=50 partial=false MAX(slots) as Slots BY full_fsgroup

 

 

 

In addition, I have another query from different index:

 

 

 

`p_flow("*",dv)` "***" reg_name=*$reg$* event_type=flow | eval fairshare = coalesce(fairshare, fsgroup) | table  fairshare | dedup  fairshare

 

 

 

 The issue is, I would like to limit the results to match only the output of the fairshare field from the second query.

Please notice that fairshare in the first query called fs_group

I've been trying many options from different past answers and nothing seem to be working. Please assist and be blessed forever.

Labels (1)
0 Karma
1 Solution

richgalloway
SplunkTrust
SplunkTrust

At the risk of suggesting something you've already tried, have you tried this?

| tstats SUM(requested_cpus) as requested_cpus, SUM(reserved_ram) as reserved_ram, SUM(requested_ram) as requested_ram, SUM(used_ram) as used_ram, SUM(compute_ram_total) as compute_ram_total, count as agg_field_seen WHERE (index=ncstat_monitor (host="$queuename$") [`p_flow("*",dv)` "***" regression_name=*$regression$* event_type=flow | eval fs_group = coalesce(fairshare, fsgroup) | fields fs_group | dedup  fs_group | format ] project=$project$ site=$site$            (slave_resource{} IN ("*")) (NOT slave_resource{} IN ("___VALUE_NONE___"))     ( host_state="normal" OR host_state="full" OR host_state="ovrld" OR host_state="sick" OR host_state="susp" OR  host_state="base" OR host_state="ready")     )
          OR (index=ncstat_monitor (host="$queuename$") fs_group=$fsgroup$ project=$project$ site=$site$ (compute_slave_res{} IN ("*")) (NOT compute_slave_res{} IN ("___VALUE_NONE___"))     ( host_state="normal" OR host_state="full" OR host_state="ovrld" OR host_state="sick" OR host_state="susp" OR  host_state="base" OR host_state="ready") ) BY _time, site, fs_group span=15min | eval query_enabled=1
...
---
If this reply helps you, Karma would be appreciated.

View solution in original post

richgalloway
SplunkTrust
SplunkTrust

At the risk of suggesting something you've already tried, have you tried this?

| tstats SUM(requested_cpus) as requested_cpus, SUM(reserved_ram) as reserved_ram, SUM(requested_ram) as requested_ram, SUM(used_ram) as used_ram, SUM(compute_ram_total) as compute_ram_total, count as agg_field_seen WHERE (index=ncstat_monitor (host="$queuename$") [`p_flow("*",dv)` "***" regression_name=*$regression$* event_type=flow | eval fs_group = coalesce(fairshare, fsgroup) | fields fs_group | dedup  fs_group | format ] project=$project$ site=$site$            (slave_resource{} IN ("*")) (NOT slave_resource{} IN ("___VALUE_NONE___"))     ( host_state="normal" OR host_state="full" OR host_state="ovrld" OR host_state="sick" OR host_state="susp" OR  host_state="base" OR host_state="ready")     )
          OR (index=ncstat_monitor (host="$queuename$") fs_group=$fsgroup$ project=$project$ site=$site$ (compute_slave_res{} IN ("*")) (NOT compute_slave_res{} IN ("___VALUE_NONE___"))     ( host_state="normal" OR host_state="full" OR host_state="ovrld" OR host_state="sick" OR host_state="susp" OR  host_state="base" OR host_state="ready") ) BY _time, site, fs_group span=15min | eval query_enabled=1
...
---
If this reply helps you, Karma would be appreciated.

yifatcy
Path Finder

Perfect, the "format" at the end did the job.

Just added [search `p_flow......... ] and it worked!

Thanks

0 Karma
Get Updates on the Splunk Community!

Splunk Observability for AI

Don’t miss out on an exciting Tech Talk on Splunk Observability for AI!Discover how Splunk’s agentic AI ...

Splunk Enterprise Security 8.x: The Essential Upgrade for Threat Detection, ...

Watch On Demand the Tech Talk, and empower your SOC to reach new heights! Duration: 1 hour  Prepare to ...

Splunk Observability as Code: From Zero to Dashboard

For the details on what Self-Service Observability and Observability as Code is, we have some awesome content ...