Uploaded image for project: 'Core Server'
  1. Core Server
  2. SERVER-17206

WT Secondaries fall behind on heavy insert workload but MMAPv1 secondaries don't

    • Type: Icon: Bug Bug
    • Resolution: Done
    • Priority: Icon: Critical - P2 Critical - P2
    • 3.0.0-rc9, 3.1.0
    • Affects Version/s: 3.0.0-rc8
    • Component/s: Replication, Storage
    • None
    • Fully Compatible
    • ALL
    • Hide

      Socialite load workload (available at https://github.com/10gen-labs/socialite).

      java -jar target/socialite-0.0.1-SNAPSHOT.jar load  --users 10000000 --maxfollows 1000 --messages 2000 --threads 32 sample-config.yml
      

      3 node replica set c3.2xlarge instances, 8 cpu, 15g.
      Replica set configuration

      {
      	"_id" : "shard1",
      	"version" : 1,
      	"members" : [
      		{
      			"_id" : 1,
      			"host" : "shard1-01.knuckleboys.com:27017"
      		},
      		{
      			"_id" : 2,
      			"host" : "shard1-02.knuckleboys.com:27017"
      		},
      		{
      			"_id" : 3,
      			"host" : "shard1-03.knuckleboys.com:27017"
      		}
      	]
      }
      

      Socialite load workload: java -jar target/socialite-0.0.1-SNAPSHOT.jar load --users 10000000 --maxfollows 1000 --messages 2000 --threads 32 sample-config.yml

      Indexes:

      [
      	{
      		"v" : 1,
      		"key" : {
      			"_id" : 1
      		},
      		"name" : "_id_",
      		"ns" : "socialite.content"
      	},
      	{
      		"v" : 1,
      		"key" : {
      			"_a" : 1,
      			"_id" : 1
      		},
      		"name" : "_a_1__id_1",
      		"ns" : "socialite.content"
      	}
      ]
      [
      	{
      		"v" : 1,
      		"key" : {
      			"_id" : 1
      		},
      		"name" : "_id_",
      		"ns" : "socialite.followers"
      	},
      	{
      		"v" : 1,
      		"unique" : true,
      		"key" : {
      			"_f" : 1,
      			"_t" : 1
      		},
      		"name" : "_f_1__t_1",
      		"ns" : "socialite.followers"
      	}
      ]
      [
      	{
      		"v" : 1,
      		"key" : {
      			"_id" : 1
      		},
      		"name" : "_id_",
      		"ns" : "socialite.following"
      	},
      	{
      		"v" : 1,
      		"unique" : true,
      		"key" : {
      			"_f" : 1,
      			"_t" : 1
      		},
      		"name" : "_f_1__t_1",
      		"ns" : "socialite.following"
      	}
      ]
      [
      	{
      		"v" : 1,
      		"key" : {
      			"_id" : 1
      		},
      		"name" : "_id_",
      		"ns" : "socialite.users"
      	}
      ]
      

      Show
      Socialite load workload (available at https://github.com/10gen-labs/socialite ). java -jar target/socialite-0.0.1-SNAPSHOT.jar load --users 10000000 --maxfollows 1000 --messages 2000 --threads 32 sample-config.yml 3 node replica set c3.2xlarge instances, 8 cpu, 15g. Replica set configuration { "_id" : "shard1" , "version" : 1, "members" : [ { "_id" : 1, "host" : "shard1-01.knuckleboys.com:27017" }, { "_id" : 2, "host" : "shard1-02.knuckleboys.com:27017" }, { "_id" : 3, "host" : "shard1-03.knuckleboys.com:27017" } ] } Socialite load workload: java -jar target/socialite-0.0.1-SNAPSHOT.jar load --users 10000000 --maxfollows 1000 --messages 2000 --threads 32 sample-config.yml Indexes: [ { "v" : 1, "key" : { "_id" : 1 }, "name" : "_id_" , "ns" : "socialite.content" }, { "v" : 1, "key" : { "_a" : 1, "_id" : 1 }, "name" : "_a_1__id_1" , "ns" : "socialite.content" } ] [ { "v" : 1, "key" : { "_id" : 1 }, "name" : "_id_" , "ns" : "socialite.followers" }, { "v" : 1, "unique" : true , "key" : { "_f" : 1, "_t" : 1 }, "name" : "_f_1__t_1" , "ns" : "socialite.followers" } ] [ { "v" : 1, "key" : { "_id" : 1 }, "name" : "_id_" , "ns" : "socialite.following" }, { "v" : 1, "unique" : true , "key" : { "_f" : 1, "_t" : 1 }, "name" : "_f_1__t_1" , "ns" : "socialite.following" } ] [ { "v" : 1, "key" : { "_id" : 1 }, "name" : "_id_" , "ns" : "socialite.users" } ] 

      Running the socialite load workload (primarily writes) against a three member replica set, 4g oplog with one secondary using MMAPv1 and the other using wiredTiger, the WT secondary starts falling behind. The MMAPv1 secondary keeps up.
      Typically, I'll see the WT secondary get to about 1500s behind the primary, then it will clear up. This take about 2 hours to reproduce. After about 8-10 hours total run time the WT secondary starts to fall behind again, then falls off the tail of the oplog.
      Screen shots of MMS during a recovered lag, and of timeline output (from https://github.com/10gen/support-tools) with correlated activity. The full timeline files are attached.

        1. socialite-3wtnode.png
          socialite-3wtnode.png
          110 kB
        2. socialite_oplog_1G.png
          socialite_oplog_1G.png
          73 kB
        3. node 3 WT secondary timeline.png
          node 3 WT secondary timeline.png
          294 kB
        4. node 2 MMAPv1 secondary timeline.png
          node 2 MMAPv1 secondary timeline.png
          213 kB
        5. node 1 WT primary timeline.png
          node 1 WT primary timeline.png
          288 kB
        6. mongod-3-.log.gz
          36 kB
        7. mongod-2-.log.gz
          35 kB
        8. mongod-1-.log.gz
          3.99 MB
        9. Dashboard___MMS__MongoDB_Management_Service.png
          Dashboard___MMS__MongoDB_Management_Service.png
          83 kB
        10. 3-ss.log
          3.32 MB
        11. 3-oplog-cs.html
          475 kB
        12. 3-op-cs.log
          1.08 MB
        13. 3lag.html
          1.21 MB
        14. 2-ss.log
          2.01 MB
        15. 2-oplog-cs.html
          63 kB
        16. 2-op-cs.log
          136 kB
        17. 2lag.html
          554 kB
        18. 1-ss.log
          3.39 MB
        19. 1-oplog-cs.html
          467 kB
        20. 1-op-cs.log
          1.06 MB
        21. 1lag.html
          1.45 MB

            Assignee:
            darren.wood@10gen.com Darren Wood
            Reporter:
            michael.grundy Michael Grundy
            Votes:
            0 Vote for this issue
            Watchers:
            11 Start watching this issue

              Created:
              Updated:
              Resolved: