Improve DDL replication related regression tests.

[anarazel]

The previous form of the test, utilizing DEBUG2, included too much
output dependent on the specifc system and version. Reformulate it to
explicitly connect to workers and show the schema there, when necessary.

@sumedhpathak I guess @samay-sharma can be the reviewer here?

[samay-sharma]

Could you check if we need the multi_alter_table_statements_0.source file?

I think the primary reason for this file was the differing DEBUG output across 9.4 / 9.5. Now that you've removed the DEBUG statements, I think it should be safe to remove the file.

[samay-sharma]

@anarazel : This looks good, other than the few minor comments I added above.

I also ran check-full and all tests pass, so should be good to go once you've looked at the above.

[anarazel]

WRT hash indexes: In theory I agree, but given that adding a gin/gist index for a similar test would be just as complicated as the hash index test (due to them only being available for few datatypes without extension), I'm inclined to leave it as is.

[onderkalaci]

@samay-sharma There are two PRs (#385 and #388) which are changing the outputs of multi_alter_table_statements.

So, are there any blockers to check-in this?

Thanks!

[onderkalaci]

Hi @anarazel, @samay-sharma

After this checked-in, the regression tests fail on PG 9.4.5 on my MacOS box on the master branch. I also tested this on an UBUNTU 14.04, the same result.

I don't understand how it passed on travis. The diff output is:

*** /Users/onderkalaci/Documents/citus_code/citus/src/test/regress/expected/multi_master_protocol.out   Thu Feb 11 19:35:17 2016
--- /Users/onderkalaci/Documents/citus_code/citus/src/test/regress/results/multi_master_protocol.out    Fri Mar 18 14:02:37 2016
***************
*** 14,20 ****
  ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
   CREATE TABLE lineitem (l_orderkey bigint NOT NULL, l_partkey integer NOT NULL, l_suppkey integer NOT NULL, l_linenumber integer NOT NULL, l_quantity numeric(15,2) NOT NULL, l_extendedprice numeric(15,2) NOT NULL, l_discount numeric(15,2) NOT NULL, l_tax numeric(15,2) NOT NULL, l_returnflag character(1) NOT NULL, l_linestatus character(1) NOT NULL, l_shipdate date NOT NULL, l_commitdate date NOT NULL, l_receiptdate date NOT NULL, l_shipinstruct character(25) NOT NULL, l_shipmode character(10) NOT NULL, l_comment character varying(44) NOT NULL)
   CREATE INDEX lineitem_time_index ON lineitem USING btree (l_shipdate)
!  ALTER TABLE public.lineitem ADD CONSTRAINT lineitem_pkey PRIMARY KEY (l_orderkey, l_linenumber)
  (3 rows)

  SELECT * FROM master_get_new_shardid();
--- 14,20 ----
  ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
   CREATE TABLE lineitem (l_orderkey bigint NOT NULL, l_partkey integer NOT NULL, l_suppkey integer NOT NULL, l_linenumber integer NOT NULL, l_quantity numeric(15,2) NOT NULL, l_extendedprice numeric(15,2) NOT NULL, l_discount numeric(15,2) NOT NULL, l_tax numeric(15,2) NOT NULL, l_returnflag character(1) NOT NULL, l_linestatus character(1) NOT NULL, l_shipdate date NOT NULL, l_commitdate date NOT NULL, l_receiptdate date NOT NULL, l_shipinstruct character(25) NOT NULL, l_shipmode character(10) NOT NULL, l_comment character varying(44) NOT NULL)
   CREATE INDEX lineitem_time_index ON lineitem USING btree (l_shipdate)
!  ALTER TABLE ONLY lineitem ADD CONSTRAINT lineitem_pkey PRIMARY KEY (l_orderkey, l_linenumber)
  (3 rows)

  SELECT * FROM master_get_new_shardid();

======================================================================

/Users/onderkalaci/Documents/citus_code/citus/src/test/regress/regression.diffs (END)

How should we tackle this?

[anarazel]

Update to a recent 9.4 release (i.e. 9.4.6), that's a difference between
the minor versions.

On Fri, Mar 18, 2016 at 5:33 AM, Önder Kalacı [email protected]
wrote:

Hi @anarazel https://github.com/anarazel, @samay-sharma
https://github.com/samay-sharma

After this checked-in, the regression tests fail on PG 9.4.5 on my MacOS
box on the master branch. I also tested this on an UBUNTU 14.04, the same
result.

I don't understand how it passed on travis. The diff output is:

*** /Users/onderkalaci/Documents/citus_code/citus/src/test/regress/expected/multi_master_protocol.out Thu Feb 11 19:35:17 2016
--- /Users/onderkalaci/Documents/citus_code/citus/src/test/regress/results/multi_master_protocol.out Fri Mar 18 14:02:37 2016

---

*** 14,20 ****

---

CREATE TABLE lineitem (l_orderkey bigint NOT NULL, l_partkey integer NOT NULL, l_suppkey integer NOT NULL, l_linenumber integer NOT NULL, l_quantity numeric(15,2) NOT NULL, l_extendedprice numeric(15,2) NOT NULL, l_discount numeric(15,2) NOT NULL, l_tax numeric(15,2) NOT NULL, l_returnflag character(1) NOT NULL, l_linestatus character(1) NOT NULL, l_shipdate date NOT NULL, l_commitdate date NOT NULL, l_receiptdate date NOT NULL, l_shipinstruct character(25) NOT NULL, l_shipmode character(10) NOT NULL, l_comment character varying(44) NOT NULL)
CREATE INDEX lineitem_time_index ON lineitem USING btree (l_shipdate)
! ALTER TABLE public.lineitem ADD CONSTRAINT lineitem_pkey PRIMARY KEY (l_orderkey, l_linenumber)
(3 rows)

SELECT * FROM master_get_new_shardid();
--- 14,20 ----

---

CREATE TABLE lineitem (l_orderkey bigint NOT NULL, l_partkey integer NOT NULL, l_suppkey integer NOT NULL, l_linenumber integer NOT NULL, l_quantity numeric(15,2) NOT NULL, l_extendedprice numeric(15,2) NOT NULL, l_discount numeric(15,2) NOT NULL, l_tax numeric(15,2) NOT NULL, l_returnflag character(1) NOT NULL, l_linestatus character(1) NOT NULL, l_shipdate date NOT NULL, l_commitdate date NOT NULL, l_receiptdate date NOT NULL, l_shipinstruct character(25) NOT NULL, l_shipmode character(10) NOT NULL, l_comment character varying(44) NOT NULL)
CREATE INDEX lineitem_time_index ON lineitem USING btree (l_shipdate)
! ALTER TABLE ONLY lineitem ADD CONSTRAINT lineitem_pkey PRIMARY KEY (l_orderkey, l_linenumber)
(3 rows)

SELECT * FROM master_get_new_shardid();

/Users/onderkalaci/Documents/citus_code/citus/src/test/regress/regression.diffs (END)

How should we tackle this?

—
You are receiving this because you were mentioned.
Reply to this email directly or view it on GitHub
#360 (comment)

[anarazel]

@onderkalaci Have you verified that the above this indeed the cause of your problem?

[onderkalaci]

Yes @anarazel, I verified that is a difference between minor versions (i.e., all tests pass on 9.4.6 whereas multi_master_protocol fails on 9.4.5 and below).