The FSM workload interleaves mapReduce commands, collection drops,
and database drops to verify server availability when drops occur
during various mapReduce phases.
The showDiskLoc() shell helper is now an alias for showRecordId(). The server still
accepts $showDiskLoc for OP_QUERY style find operations. For find command, the showDiskLoc
argument has been renamed to showRecordId.
Also renames the $diskLoc field returned in the resulting documents to $recordId. Unlike
$diskLoc, the $recordId field reports a single 64 bit integer rather than separately reporting the
high-order 32 bits and low-order 32 bits, which have MMAPv1-specific meaning of file and offset.
Includes workloads for:
explain
compact
reindex
collMod
count
distinct
touch
$where
Added to blacklists in FSM runners
Fixed the way we check for storage engines
Added two options for arbitrary function execution against cluster:
- Specified via ClusterOptions as setupFunctions to be run on the
cluster before workloads are run
- As part of setup and teardown with the cluster provided as a third
argument to these workload functions
This change removes support for the use of the "wiredtiger" name (note
lowercase t) for the Wired Tiger storage engine. From now on only the
"wiredTiger" name is supported.
Includes workloads for:
aggregation
mapReduce and group workloads
creating and dropping roles and users
update(...)
basic workloads for capped collections
findAndModify
remove(...)
dropping db/collection, renaming collections, and for server status
Updates blacklists to fsm_*js runners to work around MongoDB restrictions or known bugs.
Closes #902
Signed-off-by: Benety Goh <benety@mongodb.com>