MapR-FS to ONTAP NFS

Contributors

This section provides the detailed steps needed to move MapR-FS data into ONTAP NFS by using NetApp XCP.

  1. Provision three LUNs for each MapR node and give the LUNs ownership of all MapR nodes.

  2. During installation, choose newly added LUNs for MapR cluster disks that are used for MapR-FS.

  3. Install a MapR cluster according to the MapR 6.1 documentation.

  4. Check the basic Hadoop operations using MapReduce commands such as hadoop jar xxx.

  5. Keep customer data in MapR-FS. For example, we generated approximately a terabyte of sample data in MapR-FS by using Teragen.

  6. Configure MapR-FS as NFS export.

    1. Disable the nlockmgr service on all MapR nodes.

      root@workr-138: ~$ rpcinfo -p
         program vers proto   port  service
          100000    4   tcp    111  portmapper
          100000    3   tcp    111  portmapper
          100000    2   tcp    111  portmapper
          100000    4   udp    111  portmapper
          100000    3   udp    111  portmapper
          100000    2   udp    111  portmapper
          100003    4   tcp   2049  nfs
          100227    3   tcp   2049  nfs_acl
          100003    4   udp   2049  nfs
          100227    3   udp   2049  nfs_acl
          100021    3   udp  55270  nlockmgr
          100021    4   udp  55270  nlockmgr
          100021    3   tcp  35025  nlockmgr
          100021    4   tcp  35025  nlockmgr
          100003    3   tcp   2049  nfs
          100005    3   tcp   2049  mountd
          100005    1   tcp   2049  mountd
          100005    3   udp   2049  mountd
          100005    1   udp   2049  mountd
      root@workr-138: ~$
       
      root@workr-138: ~$ rpcinfo -d 100021 3
      root@workr-138: ~$ rpcinfo -d 100021 4
    2. Export specific folders from MapR-FS on all MapR nodes in the /opt/mapr/conf/exports file. Do not export the parent folder with different permissions when you export sub folders.

      [mapr@workr-138 ~]$ cat /opt/mapr/conf/exports
      # Sample Exports file
      # for /mapr exports
      # <Path> <exports_control>
      #access_control -> order is specific to default
      # list the hosts before specifying a default for all
      #  a.b.c.d,1.2.3.4(ro) d.e.f.g(ro) (rw)
      #  enforces ro for a.b.c.d & 1.2.3.4 and everybody else is rw
      # special path to export clusters in mapr-clusters.conf. To disable exporting,
      # comment it out. to restrict access use the exports_control
      #
      #/mapr (rw)
      #karthik
      /mapr/my.cluster.com/tmp/testnfs /maprnfs3 (rw)
      #to export only certain clusters, comment out the /mapr & uncomment.
      #/mapr/clustername (rw)
      #to export /mapr only to certain hosts (using exports_control)
      #/mapr a.b.c.d(rw),e.f.g.h(ro)
      # export /mapr/cluster1 rw to a.b.c.d & ro to e.f.g.h (denied for others)
      #/mapr/cluster1 a.b.c.d(rw),e.f.g.h(ro)
      # export /mapr/cluster2 only to e.f.g.h (denied for others)
      #/mapr/cluster2 e.f.g.h(rw)
      # export /mapr/cluster3 rw to e.f.g.h & ro to others
      #/mapr/cluster2 e.f.g.h(rw) (ro)
      #to export a certain cluster, volume or a subdirectory as an alias,
      #comment out  /mapr & uncomment
      #/mapr/clustername         /alias1 (rw)
      #/mapr/clustername/vol     /alias2 (rw)
      #/mapr/clustername/vol/dir /alias3 (rw)
      #only the alias will be visible/exposed to the nfs client not the mapr path, host options as before
      [mapr@workr-138 ~]$
  7. Refresh the MapR-FS NFS service.

    root@workr-138: tmp$ maprcli nfsmgmt refreshexports
    ERROR (22) -  You do not have a ticket to communicate with 127.0.0.1:9998. Retry after obtaining a new ticket using maprlogin
    root@workr-138: tmp$ su - mapr
    [mapr@workr-138 ~]$ maprlogin password -cluster my.cluster.com
    [Password for user 'mapr' at cluster 'my.cluster.com': ]
    MapR credentials of user 'mapr' for cluster 'my.cluster.com' are written to '/tmp/maprticket_5000'
    [mapr@workr-138 ~]$ maprcli nfsmgmt refreshexports
  8. Assign a virtual IP range to a specific server or a set of servers in the MapR cluster. Then the MapR cluster assigns an IP to a specific server for NFS data access. The IPs enable high availability, which means that, if a server or network with a particular IP experiences failure, the next IP from the range of IPs can be used for NFS access.

    Note If you would like to provide NFS access from all MapR nodes, then you can assign a set of virtual IPs to each server, and you can use the resources from each MapR node for NFS data access.

    Error: Missing Graphic Image

    Error: Missing Graphic Image

    Error: Missing Graphic Image

  9. Check the virtual IPs assigned on each MapR node and use them for NFS data access.

    root@workr-138: ~$ ip a
    1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
        link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
        inet 127.0.0.1/8 scope host lo
           valid_lft forever preferred_lft forever
        inet6 ::1/128 scope host
           valid_lft forever preferred_lft forever
    2: ens3f0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 9000 qdisc mq state UP group default qlen 1000
        link/ether 90:1b:0e:d1:5d:f9 brd ff:ff:ff:ff:ff:ff
        inet 10.63.150.138/24 brd 10.63.150.255 scope global noprefixroute ens3f0
           valid_lft forever preferred_lft forever
        inet 10.63.150.96/24 scope global secondary ens3f0:~m0
           valid_lft forever preferred_lft forever
        inet 10.63.150.97/24 scope global secondary ens3f0:~m1
           valid_lft forever preferred_lft forever
        inet6 fe80::921b:eff:fed1:5df9/64 scope link
           valid_lft forever preferred_lft forever
    3: eno1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP group default qlen 1000
        link/ether 90:1b:0e:d1:af:b4 brd ff:ff:ff:ff:ff:ff
    4: ens3f1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP group default qlen 1000
        link/ether 90:1b:0e:d1:5d:fa brd ff:ff:ff:ff:ff:ff
    5: eno2: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc mq state DOWN group default qlen 1000
        link/ether 90:1b:0e:d1:af:b5 brd ff:ff:ff:ff:ff:ff
    [root@workr-138: ~$
    [root@workr-140 ~]# ip a
    1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
        link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
        inet 127.0.0.1/8 scope host lo
           valid_lft forever preferred_lft forever
        inet6 ::1/128 scope host
           valid_lft forever preferred_lft forever
    2: ens3f0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 9000 qdisc mq state UP group default qlen 1000
        link/ether 90:1b:0e:d1:5e:03 brd ff:ff:ff:ff:ff:ff
        inet 10.63.150.140/24 brd 10.63.150.255 scope global noprefixroute ens3f0
           valid_lft forever preferred_lft forever
        inet 10.63.150.92/24 scope global secondary ens3f0:~m0
           valid_lft forever preferred_lft forever
        inet6 fe80::921b:eff:fed1:5e03/64 scope link noprefixroute
           valid_lft forever preferred_lft forever
    3: eno1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP group default qlen 1000
        link/ether 90:1b:0e:d1:af:9a brd ff:ff:ff:ff:ff:ff
    4: ens3f1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP group default qlen 1000
        link/ether 90:1b:0e:d1:5e:04 brd ff:ff:ff:ff:ff:ff
    5: eno2: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc mq state DOWN group default qlen 1000
        link/ether 90:1b:0e:d1:af:9b brd ff:ff:ff:ff:ff:ff
    [root@workr-140 ~]#
  10. Mount the NFS- exported MapR-FS using the assigned virtual IP for checking the NFS operation. However, this step is not required for data transfer using NetApp XCP.

    root@workr-138: tmp$ mount -v -t nfs 10.63.150.92:/maprnfs3 /tmp/testmount/
    mount.nfs: timeout set for Thu Dec  5 15:31:32 2019
    mount.nfs: trying text-based options 'vers=4.1,addr=10.63.150.92,clientaddr=10.63.150.138'
    mount.nfs: mount(2): Protocol not supported
    mount.nfs: trying text-based options 'vers=4.0,addr=10.63.150.92,clientaddr=10.63.150.138'
    mount.nfs: mount(2): Protocol not supported
    mount.nfs: trying text-based options 'addr=10.63.150.92'
    mount.nfs: prog 100003, trying vers=3, prot=6
    mount.nfs: trying 10.63.150.92 prog 100003 vers 3 prot TCP port 2049
    mount.nfs: prog 100005, trying vers=3, prot=17
    mount.nfs: trying 10.63.150.92 prog 100005 vers 3 prot UDP port 2049
    mount.nfs: portmap query retrying: RPC: Timed out
    mount.nfs: prog 100005, trying vers=3, prot=6
    mount.nfs: trying 10.63.150.92 prog 100005 vers 3 prot TCP port 2049
    root@workr-138: tmp$ df -h
    Filesystem              Size  Used Avail Use% Mounted on
    /dev/sda7                84G   48G   37G  57% /
    devtmpfs                126G     0  126G   0% /dev
    tmpfs                   126G     0  126G   0% /dev/shm
    tmpfs                   126G   19M  126G   1% /run
    tmpfs                   126G     0  126G   0% /sys/fs/cgroup
    /dev/sdd1               3.7T  201G  3.5T   6% /mnt/sdd1
    /dev/sda6               946M  220M  726M  24% /boot
    tmpfs                    26G     0   26G   0% /run/user/5000
    gpfs1                   7.3T  9.1G  7.3T   1% /gpfs1
    tmpfs                    26G     0   26G   0% /run/user/0
    localhost:/mapr         100G     0  100G   0% /mapr
    10.63.150.92:/maprnfs3   53T  8.4G   53T   1% /tmp/testmount
    root@workr-138: tmp$
  11. Configure NetApp XCP to transfer data from the MapR-FS NFS gateway to ONTAP NFS.

    1. Configure the catalog location for XCP.

      [root@hdp2 linux]# cat /opt/NetApp/xFiles/xcp/xcp.ini
      # Sample xcp config
      [xcp]
      #catalog =  10.63.150.51:/gpfs1
      catalog =  10.63.150.213:/nc_volume1
    2. Copy the license file to /opt/NetApp/xFiles/xcp/.

      root@workr-138: src$ cd /opt/NetApp/xFiles/xcp/
      root@workr-138: xcp$ ls -ltrha
      total 252K
      drwxr-xr-x 3 root   root     16 Apr  4  2019 ..
      -rw-r--r-- 1 root   root    105 Dec  5 19:04 xcp.ini
      drwxr-xr-x 2 root   root     59 Dec  5 19:04 .
      -rw-r--r-- 1 faiz89 faiz89  336 Dec  6 21:12 license
      -rw-r--r-- 1 root   root    192 Dec  6 21:13 host
      -rw-r--r-- 1 root   root   236K Dec 17 14:12 xcp.log
      root@workr-138: xcp$
    3. Activate XCP using the xcp activate command.

    4. Check the source for NFS export.

      [root@hdp2 linux]# ./xcp show 10.63.150.92
      XCP 1.4-17914d6; (c) 2019 NetApp, Inc.; Licensed to Karthikeyan Nagalingam [NetApp Inc] until Wed Feb  5 11:07:27 2020
      getting pmap dump from 10.63.150.92 port 111...
      getting export list from 10.63.150.92...
      sending 1 mount and 4 nfs requests to 10.63.150.92...
      == RPC Services ==
      '10.63.150.92': TCP rpc services: MNT v1/3, NFS v3/4, NFSACL v3, NLM v1/3/4, PMAP v2/3/4, STATUS v1
      '10.63.150.92': UDP rpc services: MNT v1/3, NFS v4, NFSACL v3, NLM v1/3/4, PMAP v2/3/4, STATUS v1
      == NFS Exports ==
       Mounts  Errors  Server
            1       0  10.63.150.92
           Space    Files      Space    Files
            Free     Free       Used     Used Export
        52.3 TiB    53.7B   8.36 GiB    53.7B 10.63.150.92:/maprnfs3
      == Attributes of NFS Exports ==
      drwxr-xr-x --- root root 2 2 10m51s 10.63.150.92:/maprnfs3
      1.77 KiB in (8.68 KiB/s), 3.16 KiB out (15.5 KiB/s), 0s.
      [root@hdp2 linux]#
    5. Transfer the data using XCP from multiple MapR nodes from multiple source IPs and multiple destination IPs (ONTAP LIFs).

      root@workr-138: linux$ ./xcp_yatin copy --parallel 20 10.63.150.96,10.63.150.97:/maprnfs3/tg4 10.63.150.85,10.63.150.86:/datapipeline_dataset/tg4_dest
      XCP 1.6-dev; (c) 2019 NetApp, Inc.; Licensed to Karthikeyan Nagalingam [NetApp Inc] until Wed Feb  5 11:07:27 2020
      xcp: WARNING: No index name has been specified, creating one with name: autoname_copy_2019-12-06_21.14.38.652652
      xcp: mount '10.63.150.96,10.63.150.97:/maprnfs3/tg4': WARNING: This NFS server only supports 1-second timestamp granularity. This may cause sync to fail because changes will often be undetectable.
       130 scanned, 128 giants, 3.59 GiB in (723 MiB/s), 3.60 GiB out (724 MiB/s), 5s
       130 scanned, 128 giants, 8.01 GiB in (889 MiB/s), 8.02 GiB out (890 MiB/s), 11s
       130 scanned, 128 giants, 12.6 GiB in (933 MiB/s), 12.6 GiB out (934 MiB/s), 16s
       130 scanned, 128 giants, 16.7 GiB in (830 MiB/s), 16.7 GiB out (831 MiB/s), 21s
       130 scanned, 128 giants, 21.1 GiB in (907 MiB/s), 21.1 GiB out (908 MiB/s), 26s
       130 scanned, 128 giants, 25.5 GiB in (893 MiB/s), 25.5 GiB out (894 MiB/s), 31s
       130 scanned, 128 giants, 29.6 GiB in (842 MiB/s), 29.6 GiB out (843 MiB/s), 36s
      ….
      [root@workr-140 linux]# ./xcp_yatin copy  --parallel 20 10.63.150.92:/maprnfs3/tg4_2 10.63.150.85,10.63.150.86:/datapipeline_dataset/tg4_2_dest
      XCP 1.6-dev; (c) 2019 NetApp, Inc.; Licensed to Karthikeyan Nagalingam [NetApp Inc] until Wed Feb  5 11:07:27 2020
      xcp: WARNING: No index name has been specified, creating one with name: autoname_copy_2019-12-06_21.14.24.637773
      xcp: mount '10.63.150.92:/maprnfs3/tg4_2': WARNING: This NFS server only supports 1-second timestamp granularity. This may cause sync to fail because changes will often be undetectable.
       130 scanned, 128 giants, 4.39 GiB in (896 MiB/s), 4.39 GiB out (897 MiB/s), 5s
       130 scanned, 128 giants, 9.94 GiB in (1.10 GiB/s), 9.96 GiB out (1.10 GiB/s), 10s
       130 scanned, 128 giants, 15.4 GiB in (1.09 GiB/s), 15.4 GiB out (1.09 GiB/s), 15s
       130 scanned, 128 giants, 20.1 GiB in (953 MiB/s), 20.1 GiB out (954 MiB/s), 20s
       130 scanned, 128 giants, 24.6 GiB in (928 MiB/s), 24.7 GiB out (929 MiB/s), 25s
       130 scanned, 128 giants, 29.0 GiB in (877 MiB/s), 29.0 GiB out (878 MiB/s), 31s
       130 scanned, 128 giants, 33.2 GiB in (852 MiB/s), 33.2 GiB out (853 MiB/s), 36s
       130 scanned, 128 giants, 37.8 GiB in (941 MiB/s), 37.8 GiB out (942 MiB/s), 41s
       130 scanned, 128 giants, 42.0 GiB in (860 MiB/s), 42.0 GiB out (861 MiB/s), 46s
       130 scanned, 128 giants, 46.1 GiB in (852 MiB/s), 46.2 GiB out (853 MiB/s), 51s
       130 scanned, 128 giants, 50.1 GiB in (816 MiB/s), 50.2 GiB out (817 MiB/s), 56s
       130 scanned, 128 giants, 54.1 GiB in (819 MiB/s), 54.2 GiB out (820 MiB/s), 1m1s
       130 scanned, 128 giants, 58.5 GiB in (897 MiB/s), 58.6 GiB out (898 MiB/s), 1m6s
       130 scanned, 128 giants, 62.9 GiB in (900 MiB/s), 63.0 GiB out (901 MiB/s), 1m11s
       130 scanned, 128 giants, 67.2 GiB in (876 MiB/s), 67.2 GiB out (877 MiB/s), 1m16s
    6. Check the load distribution on the storage controller.

      Hadoop-AFF8080::*> statistics show-periodic -interval 2 -iterations 0 -summary true -object nic_common -counter rx_bytes|tx_bytes -node Hadoop-AFF8080-01 -instance e3b
      Hadoop-AFF8080: nic_common.e3b: 12/6/2019 15:55:04
       rx_bytes tx_bytes
       -------- --------
          879MB   4.67MB
          856MB   4.46MB
          973MB   5.66MB
          986MB   5.88MB
          945MB   5.30MB
          920MB   4.92MB
          894MB   4.76MB
          902MB   4.79MB
          886MB   4.68MB
          892MB   4.78MB
          908MB   4.96MB
          905MB   4.85MB
          899MB   4.83MB
      Hadoop-AFF8080::*> statistics show-periodic -interval 2 -iterations 0 -summary true -object nic_common -counter rx_bytes|tx_bytes -node Hadoop-AFF8080-01 -instance e9b
      Hadoop-AFF8080: nic_common.e9b: 12/6/2019 15:55:07
       rx_bytes tx_bytes
       -------- --------
          950MB   4.93MB
          991MB   5.84MB
          959MB   5.63MB
          914MB   5.06MB
          903MB   4.81MB
          899MB   4.73MB
          892MB   4.71MB
          890MB   4.72MB
          905MB   4.86MB
          902MB   4.90MB