BUG #7546: Backups on hot standby cancelled despite hot_standby=on

The following bug has been logged on the website:

Bug reference: 7546
Logged by: Stuart Bishop
Email address: stuart(at)stuartbishop(dot)net
PostgreSQL version: 9.1.5
Operating system: Ubuntu 12.10
Description:

I have a primary and a hot standby using streaming replication. The hot
standby specifies 'hot_standby_feedback=on' with other replication settings
set to default.

If a vacuum occurs on the primary while pg_dump is dumping a large table,
the pg_dump is cancelled, usually with the following error:

ERROR: canceling statement due to conflict with recovery
DETAIL: User was holding shared buffer pin for too long.

I can excercise this problem with the following script:

#!/bin/sh

dbname="repl_test"
master_port=5432
slave_port=5434

rows=2000000

slow="pv --rate-limit 20k"

createdb -p $master_port $dbname

psql -p $master_port -d $dbname -f - <<EOM
CREATE TABLE IF NOT EXISTS BigStuff (
a serial primary key, b integer, c text, d text)
WITH (autovacuum_enabled = FALSE);

INSERT INTO BigStuff (b, c, d)
SELECT i,md5(i::text),reverse(md5(i::text))
FROM generate_series(1,${rows}) AS i;

DELETE FROM BigStuff WHERE random() < 0.15;
EOM

synced=0
while [ $synced -ne 1 ]
do sleep 5
synced=`psql -qtA -p $master_port -d $dbname -c "SELECT
(pg_current_xlog_location() = min(replay_location))::integer FROM
pg_stat_replication;"`
echo synced $synced synced
done

(pg_dump -p $slave_port --format=c $dbname | $slow > /dev/null) &

sleep 5;

psql -p $master_port -d $dbname -c "vacuum verbose BigStuff;"