[check_postgres] same_schema check is failing on 2.15.4, but worked fine on 2.14.3

[Aleksey Tsalolikhin]

Hi, Any advice?  Am I doing something wrong, or is this a bug in
same_schema check?

Thanks,
Aleksey

On Tue, Jan 4, 2011 at 12:43 PM, Aleksey Tsalolikhin
<atsaloli.tech at gmail.com> wrote:
> I have a couple of Slony replication sets, and use same_schema action
> to check against corruption on the slave.
>
> My Slony replication sets have "_cluster" in the name, so I filter out
> the Slony bits as follows:
>
> /usr/bin/check_postgres.pl-2.14.3 --verbose --action same_schema
> -dbhost=db1 --dbhost2=db2 --dbname=mydbname --warning='noposition
> noschema=_cluster notrigger=_cluster nofunction=_cluster'
>
> POSTGRES_SAME_SCHEMA OK: DB "mydbname" (host:db1 => db22) Both
> databases have identical items | time=0.01
>
>
> However, 2.15.4 reports the same_schema check as CRITICAL:
>
>
> POSTGRES_SAME_SCHEMA CRITICAL: DB "mydbname" (host:db1 => db2)
> Databases were different. Items not matched: 21 | time=0.01
> Index on 1 but not 2:
> "_mydbname_app00_cluster"."PartInd_mydbname_app00_cluster_sl_log_1-node-1"
> ON
>  Index on 1 but not 2:
> "_mydbname_app00_cluster"."PartInd_mydbname_app00_cluster_sl_log_2-node-1"
> ON
>  Index on 1 but not 2: "_mydbname_app00_cluster"."sl_event-pkey" ON
>  Index on 1 but not 2: "_mydbname_app00_cluster"."sl_listen-pkey" ON
>  Index on 1 but not 2: "_mydbname_app00_cluster"."sl_node-pkey" ON
>  Index on 1 but not 2: "_mydbname_app00_cluster"."sl_nodelock-pkey" ON
>  Index on 1 but not 2: "_mydbname_app00_cluster"."sl_path-pkey" ON
>  Index on 1 but not 2: "_mydbname_app00_cluster"."sl_sequence-pkey" ON
>  Index on 1 but not 2: "_mydbname_app00_cluster"."sl_set-pkey" ON
>  Index on 1 but not 2: "_mydbname_app00_cluster"."sl_setsync-pkey" ON
>  Index on 1 but not 2: "_mydbname_app00_cluster"."sl_subscribe-pkey" ON
>  Index on 1 but not 2: "_mydbname_app00_cluster"."sl_table-pkey" ON
>  Index on 1 but not 2: "_mydbname_app00_cluster"."sl_trigger-pkey" ON
>  Index on 1 but not 2: "_mydbname_app00_cluster".sl_confirm_idx1 ON
>  Index on 1 but not 2: "_mydbname_app00_cluster".sl_confirm_idx2 ON
>  Index on 1 but not 2: "_mydbname_app00_cluster".sl_log_1_idx1 ON
>  Index on 1 but not 2: "_mydbname_app00_cluster".sl_log_2_idx1 ON
>  Index on 1 but not 2: "_mydbname_app00_cluster".sl_registry_pkey ON
>  Index on 1 but not 2: "_mydbname_app00_cluster".sl_seqlog_idx ON
>  Index on 1 but not 2: "_mydbname_app00_cluster".sl_sequence_seq_reloid_key ON
>  Index on 1 but not 2: "_mydbname_app00_cluster".sl_table_tab_reloid_key ON
>
> I've tried adding "noindex=_cluster" to the warning but I still get
> the same output.
>
> I've had to fall back on 2.14.3 for the same_schema check for now, but
> could you please
> help me get this working with 2.15.4?
>
> Best,
> -at
>