Installation

Hitachi monitoring: VSP G, VSP F, HCP, HNAS, HUS-VM, HUS, AMS


In case of usage of Virtual Appliance
  • Use local account lpar2rrd for hosting of STOR2RRD on the virtual appliance
  • Use /home/stor2rrd/stor2rrd as the product home
  • Use lpar2rrd account on storage as it is configured in /home/stor2rrd/stor2rrd/etc/stor2rrd.cfg (STORAGE_USER)
The program uses Hitachi Storage Navigator Modular 2 CLI interface (HSNM2 CLI).
The CLI is included in the same CD from where you got the SNM UI. If you dont have the CD you should ask HDS for the SNM CD (usually distributed as an ISO image).

Install HUS HSNM2 CLI

  • Allow access from the STOR2RRD host to the both HUS storage controllers on port 2000 (28355 when using secure communication).
    Test if port is open:
    $ perl /home/stor2rrd/stor2rrd/bin/conntest.pl 192.168.1.1 2000
      Connection to "192.168.1.1" on port "2000" is ok
    
  • HUS HSNM2 CLI apparently needs libstdc++ package. Assure that it is installed.
    $ rpm -qa|grep libstdc++
      libstdc++-4.8.3-9.el7.x86_64
    
    Install it if it is not listed:
    # yum install libstdc++     # (RedHat, prp. Debian/Ubuntu )
    # apt-get install libstdc++ # (Debian/Ubuntu )
    
  • Get HUS HSNM2 CLI package for your operating system and install it under root:
    # mkdir /usr/stonavm/
    # cd /usr/stonavm/
    # tar xf /tmp/HSNM2-2810-A-CLI-P01.tar
    # chown stor2rrd /usr/stonavm   # This must be owned by stor2rrd (lpar2rrd on the virtual appliance) !!!!
    
  • Set environment in your actual shell (just copy&paste to the cmd line)
    LIBPATH=/usr/stonavm/lib:$LIBPATH
    SHLIB_PATH=/usr/stonavm/lib:$SHLIB_PATH
    LD_LIBRARY_PATH=/usr/stonavm/lib:$LD_LIBRARY_PATH
    STONAVM_HOME=/usr/stonavm
    STONAVM_ACT=on
    STONAVM_RSP_PASS=on
    PATH=$PATH:/usr/stonavm
    export  LIBPATH SHLIB_PATH LD_LIBRARY_PATH STONAVM_HOME STONAVM_ACT STONAVM_RSP_PASS PATH
    
  • Register the storage systems to be monitored (adjust IPs in below examples)
    • Automatically
      # auunitaddauto -ip 192.168.1.1 192.168.1.2
      
    • Manuallly
      # auunitadd -unit HUS110 -LAN -ctl0 192.168.1.1 -ctl1 192.168.1.2
      
    • Using secure communication (port 28355 on storage):
      # auunitadd -unit HUS110 -LAN -ctl0 192.168.1.1 -ctl1 192.168.1.2 -communicationtype secure
      
  • Create user stor2rrd (lpar2rrd on the Appliance) on the storage with role Storage Administrator (View Only)

  • Register user access for the storage
    # auaccountenv -set -uid stor2rrd -authentication 
    
  • Test connectivity
    # auunitref
    # auunitinfo -unit HUS110
    

STOR2RRD storage configuration

  • All actions below under stor2rrd user (lpar2rrd on Virtual Appliance)

  • Add storage into configuration from the UI:
    Settings icon ➡ Storage ➡ New ➡ Vendor:device ➡ Hitachi HUS
    HDS HUS Storage management

  • Make sure you have enough of disk space on the filesystem where is STOR2RRD installed
    Roughly you might count 2 - 30 GB per a storage (it depends on number of volumes, 30GB for 5000 volumes)
    $ df -g /home   # AIX
    $ df -h /home   # Linux
    
  • Test storage connectivity, either from the UI or from the command line:
    $ cd /home/stor2rrd/stor2rrd
    $ ./bin/config_check.sh 
      =========================
      STORAGE: HUS110: HUS 
      =========================
      /usr/stonavm/auunitinfo -unit HUS110
      connection ok
    
  • Schedule to run storage agent from stor2rrd crontab (lpar2rrd on Virtual Appliance, it might already exist there)
    $ crontab -l | grep load_husperf.sh
    $
    
    Add if it does not exist as above
    $ crontab -e
    
    # Hitachi HUS && AMS 2000 storage agent 
    0,5,10,15,20,25,30,35,40,45,50,55 * * * * /home/stor2rrd/stor2rrd/load_husperf.sh > /home/stor2rrd/stor2rrd/load_husperf.out 2>&1
    
    Assure there is already an entry with the UI creation running once an hour in crontab
    $ crontab -e
    
    # STOR2RRD UI (just ONE entry of load.sh must be there)
    5 * * * * /home/stor2rrd/stor2rrd/load.sh > /home/stor2rrd/stor2rrd/load.out 2>&1
    
  • Let run the storage agent for 15 - 20 minutes to get data, then:
    $ cd /home/stor2rrd/stor2rrd
    $ ./load.sh
    
  • Go to the web UI: http://<your web server>/stor2rrd/
    Use Ctrl-F5 to refresh the web browser cache.
In case of usage of Virtual Appliance
  • Use local account lpar2rrd for hosting of STOR2RRD on the virtual appliance
  • Use /home/stor2rrd/stor2rrd as the product home

Hitachi VSP G/F/E, HUS-VM monitoring setups

  1. REST API, Export Tool: use it on new models which support it, it has much easier setup
  2. CCI, SNMP, Export Tool: it works on all models
  3. Hitachi Configuration Manager REST API / Export Tool: since STOR2RRD v7.30
Hitachi VSP monitoring scenarios

1. Hitachi VSPG REST API, Export Tool

Supported models

Prerequisites

  • Allow communication from STOR2RRD host to all Hitachi VSPG storage SVP IP and nodes on TCP port 1099.
    Nodes IP, allow TCP port 443.
    At least VSP 5500 has not available Nodes, then use SVP IP instead and allow also TCP port 443.
    $ perl /home/stor2rrd/stor2rrd/bin/conntest.pl 192.168.1.1 1099
      Connection to "192.168.1.3" on port "1099" is ok
    $ perl /home/stor2rrd/stor2rrd/bin/conntest.pl 192.168.1.1 443
      Connection to "192.168.1.3" on port "443" is ok
    
  • Storage configuration:
    Create user stor2rrd on the storage, read only access
    Do not use shell special characters like #!?|$*[]\{}`"'& in the password, use ;:.+-%@ instead.
    You can also follow this docu to fully prepare storage for monitoring

Installation of Hitachi Export Tool

    It is typically located on a CD that comes packaged with the Service Processor on the HDS USP Array. The Export Tool can also be obtained by contacting HDS support. (CD location: /ToolsPack/ExportTool)
    Hitachi produces a new Export Tool for each release of the firmware. So unless you make sure all of the storage are running on the same firmware version then you will need to obtain the appropriate version of the Export Tool to meet the firmware version you are running at the site.

    Find our firmware release of your storage (like 83-01-28/00).
    Export Tool version must match the SVP firmware version.
    Install each version of the Export Tool into separate directory named as firmware of your storage (just 6 numbers like in this example firmware 83-01-28) under root user:
    # mkdir /opt/hds
    # mkdir /opt/hds/83-01-28
    # cd /opt/hds/83-01-28
    # tar xvf export-tool.tar
    # chmod 755 runUnix.sh runUnix.bat # note one of these files exists only
    # chown -R stor2rrd /opt/hds
    # chown -R lpar2rrd /opt/hds  # do this on the Virtual Appliance where is all under "lpar2rrd" user
    
    Might work that higher Export Tool version works even with lower storage firmware like in that example (Export Tool 83-01-28 and storage firmware 73-03-57).
    In this case you do not need to install older Export Tool, just make a symlink.
    # cd /opt/hds/
    # ls
      83-01-28
    # ln -s 83-01-28 73-03-57
    
    Test Export Tool 2
    $ cd /opt/hds/<firmware level> # example /opt/hds/88-03-23
    $ sh ./runUnix.bat show interval -ip <ip controller> -login <user> <password>
      Interval : 5 min
      show interval command success
    

    Directory /opt/hds is optional, it is configurable in /home/stor2rrd/stor2rrd/etc/stor2rrd.cfg : VSP_CLIDIR=/opt/hds
    The HDS Performance Monitor License must exist for each array and monitoring must be enabled.
    Storage configuration example
    Java 11 is required by the Export Tool for new VSPG 5500 boxes.

Allow monitoring of CU and WWN

    Note this configuration option do not have to be in place on all modes or firmwares, you might ignore it if you do not find it on your storage.
  • CU
    Hitachi CU menu

    Hitachi CU selection

  • WWN
    Note that monitoring apparently cannot be enabled when WWNs per Port exceeds the maximum of 32.
    In this case you will not have direct per host data but host data will be aggregated from attached volumes (it might mislead when volumes have attached more hosts).

    Hitachi WWN menu

    Hitachi WWN selection


    When you still do not have data then re-enable of monitoring might help. Hitachi WWN

STOR2RRD storage configuration

  • Configure Export Tool installation directory in /home/stor2rrd/stor2rrd/etc/stor2rrd.cfg when is in different pathen then /opt/hds
     VSP_CLIDIR=/opt/hds
    
  • Add storage into configuration from the UI:
    Settings icon ➡ Storage ➡ New ➡ Vendor:device ➡ Hitachi VSP-G ➡ REST API
    When your storage model has only SVP, then use just SVP IP/hostname for SVP field and same for Node1
    When your storage does not have SVP, then use for the SVP field Node1 IP/hostname
    HDS VSP Storage management

  • All actions below under stor2rrd user on STOR2RRD server (lpar2rrd on Virtual Appliance)

  • Assure you have enough of disk space on the filesystem where is STOR2RRD installed
    Roughly you might count 2 - 30 GB per a storage (it depends on number of volumes, 30GB for 5000 volumes)
    $ df -g /home   # AIX
    $ df -h /home   # Linux
    
  • Schedule to run storage agent from stor2rrd crontab (lpar2rrd on Virtual Appliance, it might already exist there)
    $ crontab -l | grep load_vspgperf.sh
    $
    
    Add if it does not exist as above
    $ crontab -e
    
    # Hitachi VSP-G
    0,5,10,15,20,25,30,35,40,45,50,55 * * * * /home/stor2rrd/stor2rrd/load_vspgperf.sh > /home/stor2rrd/stor2rrd/load_vspgperf.out 2>&1
    
    Assure there is already an entry with the UI creation running once an hour in crontab
    $ crontab -e
    
    # STOR2RRD UI (just ONE entry of load.sh must be there)
    5 * * * * /home/stor2rrd/stor2rrd/load.sh > /home/stor2rrd/stor2rrd/load.out 2>&1
    
  • Let run the storage agent for 15 - 20 minutes to get data, then:
    $ cd /home/stor2rrd/stor2rrd
    $ ./load.sh
    
  • Go to the web UI: http://<your web server>/stor2rrd/
    Use Ctrl-F5 to refresh the web browser cache.


2. Hitachi VSPG CCI, SNMP, Export Tool

The program uses 3 Hitachi APIs. You have to install and configure all of them.
  • Command Control Interface (CCI)
  • Hitachi Export tool
  • SNMP API to get Health status
You might also look into very detaily described installation procedure on www.sysaix.com.
Note that Virtual Storage Machines (VSM) feature is not supported by the tool.

Storage configuration

Installation of Hitachi CCI

  • Allow communication from STOR2RRD host to all Hitachi VSPG storage SVP IP on TCP ports 1099, 51100
    Note new firmwares might use 51101 or 51099 ports instead of 51100.
    Test open ports for TCP protocols to SVP IP:
    $ perl /home/stor2rrd/stor2rrd/bin/conntest.pl 192.168.1.1 1099
      Connection to "192.168.1.3" on port "1099" is ok
    $ perl /home/stor2rrd/stor2rrd/bin/conntest.pl 192.168.1.1 51100
      Connection to "192.168.1.3" on port "51100" is ok
    
    Allow communication from STOR2RRD host to all storage node IP on UDP 31001
    How to test UDP

  • Create user stor2rrd on the storage, read only access
    Do not use shell special characters like #!?|$*[]\{}`"'& in the password, use ;:.+-%@ instead.

  • Obtain CCI installation package from your Hitachi representatives.
    • Install it from .iso image under root account
      Mount ISO image:
      • AIX command:
        # loopmount -i /HS042_77.iso -o "-V cdrfs -o ro" -m /mnt
        
      • Linux (Virtual Appliance) command:
        # mount -o loop,ro HS042_77.iso /mnt 
        
        Assure you have installed 32 bits libraries, install them if do not
        # yum -y install glibc.i686
        
    • Create target directory:
      # mkdir /etc/HORCM
      # cd /mnt
      # ./RMinstsh
      
    • Install from CD
      # mkdir /opt
      # cpio -idmu < /dev/XXXX    # where XXXX = I/O device with install media
      # ln -s /opt/HORCM /HORCM
      
    • Execute the CCI installation command:
      # /HORCM/horcminstall.sh
      
    • Verify installation of the proper version using the raidqry command:
      # raidqry -h
        Model: RAID-Manager/HP-UX
        Ver&Rev: 01-29-03/05
        Usage: raidqry [options]
      
    • Assure that everything is executable and writeable by stor2rrd user
      Use lpar2rrd user on the Virtual Appliance.
      This is a must! Under root identification execute this:
      # touch /HORCM/etc/USE_OLD_IOCT 
      # chown stor2rrd /HORCM
      # chown -R stor2rrd /HORCM/* /HORCM/.uds 
      # chown -R lpar2rrd /HORCM/* /HORCM/.uds  # do this on the Virtual Appliance where is all under "lpar2rrd" user
      # chmod 755 /HORCM /HORCM/usr/bin /HORCM/usr/bin/* /HORCM/log* /HORCM/etc/horcmgr /HORCM/etc/*conf /HORCM/.uds/ 
      

Configuration of CCI

  • CCI communication with storage can be done either via LAN (it is something described below) or via command device (SAN attached volume from the storage).
    When you have many storage in place 40+ then use rather command device as LAN communication might not be reliable enough. CCI command device configuration procedure

  • Each storage must have its own config file /etc/horcmXX.conf

  • Check if local ports 10001 and 10002 are not used (nothing is listening there)
    # netstat -an|grep -i listen| egrep "11001|11002"
    
  • storage with controllers IP 192.168.1.1 and 192.168.1.2, conf file /etc/horcm1.conf will use local port 11001 (UDP)
    Use storage node IP. SVP IP must be used in etc/storage-list.cfg further.
    # vi /etc/horcm1.conf
    
    HORCM_MON
    # ip_address service poll(10ms) timeout(10ms)
    localhost    11001   1000       3000
    HORCM_CMD
    # dev_name dev_name dev_name
    \\.\IPCMD-192.168.1.1-31001   \\.\IPCMD-192.168.1.2-31001
    
  • storage with IP 192.168.1.10 and 192.168.1.11, conf file /etc/horcm2.conf
    change localhost port to 11002 (from 11001 which is used above)
    # vi /etc/horcm2.conf
    
    HORCM_MON
    # ip_address service poll(10ms) timeout(10ms)
    localhost    11002   1000       3000
    HORCM_CMD
    # dev_name dev_name dev_name
    \\.\IPCMD-192.168.1.10-31001  \\.\IPCMD-192.168.1.11-31001
    
  • Start it under stor2rrd account (definitely not under root!). Use lpar2rrd account on the Virtual Appliance
    This starts HORM instance 1 (/etc/horcm1.conf)
    # su - stor2rrd
    $ /HORCM/usr/bin/horcmstart.sh 1
    
  • Start HORM instance 1 & 2 (/etc/horcm1.conf & /etc/horcm2.conf)
    # su - stor2rrd
    $ /HORCM/usr/bin/horcmstart.sh 1 2
    
  • Check if they are running
    $ ps -ef | grep horcm
      stor2rrd 19660912 1 0 Feb 26 - 0:03 horcmd_02
      stor2rrd 27590770 1 0 Feb 26 - 0:09 horcmd_01
    
  • Place it into operating system start/stop scripts
    # su - stor2rrd -c "/HORCM/usr/bin/horcmstart.sh 1 2"
    # su - stor2rrd -c "/HORCM/usr/bin/horcmshutdown.sh 1 2"
    
  • When HORCM does not want to start then
    • Assure that filesystem permission are fine for /HORCM (owned by stor2rrd user)
    • Check if connection to storage IP nodes is allowed: how to test UDP

Installation of Hitachi Export Tool

    It is typically located on a CD that comes packaged with the Service Processor on the HDS USP Array. The Export Tool can also be obtained by contacting HDS support. (CD location: /ToolsPack/ExportTool)
    Hitachi produces a new Export Tool for each release of the firmware. So unless you make sure all of the storage are running on the same firmware version then you will need to obtain the appropriate version of the Export Tool to meet the firmware version you are running at the site.

    Find our firmware release of your storage (83-01-28/00) identified by /etc/horcm1.conf (-I1):
    Export Tool version must match the SVP firmware version.
    Under stor2rrd account!
    # su - stor2rrd
    $ raidcom -login stor2rrd <password> -I1
    $ raidqry -l -I1 
      No  Group    Hostname     HORCM_ver   Uid   Serial#   Micro_ver     Cache(MB)
       1    ---   localhost   01-35-03-08     0   471234    83-01-28/00      320000
    $ raidcom -logout -I1
    
    Install each version of the Export Tool into separate directory named as firmware of your storage (just 6 numbers like in this example firmware 83-01-28) under root user:
    # mkdir /opt/hds
    # mkdir /opt/hds/83-01-28
    # cd /opt/hds/83-01-28
    # tar xvf export-tool.tar
    # chmod 755 runUnix.sh runUnix.bat # note one of these files exists only
    # chown -R stor2rrd /opt/hds
    # chown -R lpar2rrd /opt/hds  # do this on the Virtual Appliance where is all under "lpar2rrd" user
    
    Might work that higher Export Tool version works even with lower storage firmware like in that example (Export Tool 83-01-28 and storage firmware 73-03-57).
    In this case you do not need to install older Export Tool, just make a symlink.
    # cd /opt/hds/
    # ls
      83-01-28
    # ln -s 83-01-28 73-03-57
    
    Test Export Tool 2
    $ cd /opt/hds/<firmware level> # example /opt/hds/88-03-23
    $ sh ./runUnix.bat show interval -ip <ip controller> -login <user> <password>
      Interval : 5 min
      show interval command success
    

    Directory /opt/hds is optional, it is configurable in /home/stor2rrd/stor2rrd/etc/stor2rrd.cfg : VSP_CLIDIR=/opt/hds
    The HDS Performance Monitor License must exist for each array and monitoring must be enabled.
    Storage configuration example

Allow monitoring of CU and WWN

    Note this configuration option do not have to be in place on all modes or firmwares, you might ignore it if you do not find it on your storage.
  • CU
    Hitachi CU menu

    Hitachi CU selection

  • WWN
    Note that monitoring apparently cannot be enabled when WWNs per Port exceeds the maximum of 32.
    In this case you will not have direct per host data but host data will be aggregated from attached volumes (it might mislead when volumes have attached more hosts).

    Hitachi WWN menu

    Hitachi WWN selection


    When you still do not have data then re-enable of monitoring might help. Hitachi WWN

Health status

    Only the way to get health status from older storage models is SNMP protocol.

    Install snmpwalk

    Skip that in case you are on our Virtual Appliance
    • AIX
      Download Net-SNMP packages and install them.
      Do not use the latest packages on AIX, it does not work, use net-snmp-5.6.2.1-1!
      # umask 0022
      # rpm -Uvh net-snmp-5.6.2.1-1 net-snmp-utils-5.6.2.1-1 net-snmp-perl-5.6.2.1-1
      
      Make sure
      • you use PERL=/opt/freeware/bin/perl in etc/stor2rrd.cfg
      • PERL5LIB in etc/stor2rrd.cfg contains /opt/freeware/lib/perl5/vendor_perl/5.8.8/ppc-thread-multi path
    • Linux
      Follow this to get installed SNMP support in the tool on RHEL8 and CentOS8.
      # umask 0022
      # yum install net-snmp
      # yum install net-snmp-utils
      # yum install net-snmp-perl
      
      Note you might need to allow optional repositories on RHEL to yum can find it
      # subscription-manager repos --list
      ...
      # subscription-manager repos --enable rhel-7-server-optional-rpms
      
      Use rhel-7-for-power-le-optional-rpms for Linux on Power etc ...

    • Linux Debian/Ubuntu
      % umask 0022
      % apt-get install snmp libsnmp-perl snmp-mibs-downloader
      
      Assure that this line is commented out in /etc/snmp/snmp.conf
      #mibs :
      
      If apt-get does not find snmp-mibs-downloader package then enable contrib and non-free repositories.

    Storage configuration - VSP-G

    Allow SNMP on the storage, configure protocol (SNMP version), community string and permit STOR2RRD IP/hostname.
       Hitachi VSPG SNMP setup


    Storage configuration - HUS-VM

    1. Allow SNMP on the storage, below example uses SNMP v1
    2. Configure community string, example has "Public" set, note the first letter is uppercase
    3. Permit STOR2RRD IP/hostname

       Hitachi HUS-VM SNMP setup


    Network communication

    Allow comunication between STOR2RRD server and the storage on CNTL IP and port 161 UDP
    You can test network visibility through this test:
    perl /home/stor2rrd/stor2rrd/bin/conntest_udp.pl vspg_CNTL_host.example.com 161
      UDP connection to "vspg_CNTL_host.example.com" on port "161" is ok
    

3. Hitachi Configuration Manager, Export Tool

Hitachi Configuration Manager can be used as a replacement of Hitachi CCI, its configuration is much easier than CCI.
Either use direct storage REST API instead of CCI, which is not however available on older systems like the VSP G1000 or VSP G200/400 which have no REST API support.
Therefore its implemented Hitachi Configuration Manager support to enable REST API access even for these storage devices.

Configuration is easy, create an account on Hitachi Configuration Manager and put its hostname and user credentials into STOR2RRD configuration.
You have to define also device id of the storage.
Hitachi Configuration Manager is apparently free for Hitachi customers.

Prerequisities

  • Open port to the HCP are 23450 (HTTP) or 23451 (HTTPS) from STOR2RRD server
  • It uses same user account which is used for Export Tool already created on the back-end storage

STOR2RRD storage configuration

  • All actions below under stor2rrd user (lpar2rrd on Virtual Appliance)

  • Configure Export Tool installation directory in /home/stor2rrd/stor2rrd/etc/stor2rrd.cfg
     VSP_CLIDIR=/opt/hds
    
  • Add storage into configuration from the UI:
    Settings icon ➡ Storage ➡ New ➡ Vendor:device ➡ Hitachi VSP-G
    Hitachi Configuration Manager - Storage management

  • Read this for setting up of health status monitoring via SNMP.

  • Make sure you have enough of disk space on the filesystem where is STOR2RRD installed
    Roughly you might count 2 - 30 GB per a storage (it depends on number of volumes, 30GB for 5000 volumes)
    $ df -g /home   # AIX
    $ df -h /home   # Linux
    
  • Test storage connectivity, either from the UI or from the command line:
    $ cd /home/stor2rrd/stor2rrd
    $ ./bin/config_check.sh 
      =========================
      STORAGE: VSPG-600 : VSPG
      =========================
      connection ok
    
  • Schedule to run storage agent from stor2rrd crontab (lpar2rrd on Virtual Appliance, it might already exist there)
    $ crontab -l | grep load_vspgperf.sh
    $
    
    Add if it does not exist as above
    $ crontab -e
    
    # Hitachi VSP-G
    0,5,10,15,20,25,30,35,40,45,50,55 * * * * /home/stor2rrd/stor2rrd/load_vspgperf.sh > /home/stor2rrd/stor2rrd/load_vspgperf.out 2>&1
    
    Assure there is already an entry with the UI creation running once an hour in crontab
    $ crontab -e
    
    # STOR2RRD UI (just ONE entry of load.sh must be there)
    5 * * * * /home/stor2rrd/stor2rrd/load.sh > /home/stor2rrd/stor2rrd/load.out 2>&1
    
  • Let run the storage agent for 15 - 20 minutes to get data, then:
    $ cd /home/stor2rrd/stor2rrd
    $ ./load.sh
    
  • Go to the web UI: http://<your web server>/stor2rrd/
    Use Ctrl-F5 to refresh the web browser cache.
In case of usage of Virtual Appliance
  • Use local account lpar2rrd for hosting of STOR2RRD on the virtual appliance
  • Use /home/stor2rrd/stor2rrd as the product home
The tool uses SNMP protocol to get performance and configuration data from the HNAS storage.

There is apparently two HW different HNAS products.
  • Block storage system with a separate NAS head in front ➡ use EVS IP in connection details
  • Unfied system which is a blockstorage with an integrated NAS head ➡ connect it to one of the nodes.

Install Prerequisites (skip that in case of Virtual Appliance)

  • AIX
    Download Net-SNMP packages and install them.
    Do not use the latest packages on AIX, it does not work, use net-snmp-5.6.2.1-1!
    # umask 0022
    # rpm -Uvh net-snmp-5.6.2.1-1 net-snmp-utils-5.6.2.1-1 net-snmp-perl-5.6.2.1-1
    
    Make sure
    • you use PERL=/opt/freeware/bin/perl in etc/stor2rrd.cfg
    • PERL5LIB in etc/stor2rrd.cfg contains /opt/freeware/lib/perl5/vendor_perl/5.8.8/ppc-thread-multi path

  • Linux
    # umask 0022
    # yum install net-snmp
    # yum install net-snmp-utils
    # yum install net-snmp-perl
    
    Note you might need to allow optional repositories on RHEL to yum can find it
    # subscription-manager repos --list
    ...
    # subscription-manager repos --enable rhel-7-server-optional-rpms
    
    Use rhel-7-for-power-le-optional-rpms for Linux on Power etc ...

  • Linux Debian/Ubuntu
    % umask 0022
    % apt-get install snmp libsnmp-perl snmp-mibs-downloader
    
    Assure that this line is commented out in /etc/snmp/snmp.conf
    #mibs :
    
    If apt-get does not find snmp-mibs-downloader package then enable contrib and non-free repositories.

Enable SNMP v2c or v3 protocol on the storage

  • SNMP v2c
    Hitachi docu, follow section "Configuring SNMP access"

    Navigate to: Home ➡ Server ➡ Settings ➡ SNMP Access Configuration
    Select SNMP v2c, leave port 161, add stor2rrd server as alllowed host and set community string
    Note: do not use SNMP v1, it does not work

  • SNMP v3
    Hitachi docu, follow section "Configuring SNMPv3 access"

    Use the CLI command snmp-protocol to configure SNMPv3.
    When SNMPv3 is enabled the SNMP agent will not respond to SNMPv1 or SNMPv2c requests.
    $ ssh supervisor@<EVS hostname/IP>
    
    HNAS1:$ snmp-protocol -v v3
    HNAS1:$ snmp-protocol
            Protocol:      SNMPv3               
    
    Add users with the snmpv3-user-add command.
    HNAS1:$ snmpv3-user-add stor2rrd 
            Please enter the authentication password:     ********
            Please re-enter the authentication password:  ********
            Please enter the privacy password:    ********
            Please re-enter the privacy password: ********
    

Enable enhanced stats

  • It must be anabled otherwise Pool and Disk stats will not work
    Use HNAS CLI (ssh) to enable it under supervisor user:
    $ ssh supervisor@<EVS hostname/IP>
    
    HNAS1:$ enhanced-perf-stats
      Enhanced stats gathering: disabled
    HNAS1:$ enhanced-perf-stats -enable
    

Allow network access

  • Allow access from the STOR2RRD host to the storage (EVS IP) on port 161 UDP .
    Test if port is open:
    $ perl /home/stor2rrd/stor2rrd/bin/conntest_udp.pl 192.168.1.1 161
      Connection to "192.168.1.1" on port "161" is ok
    

STOR2RRD storage configuration

  • All actions below under stor2rrd user (lpar2rrd on Virtual Appliance)

  • Add storage into configuration from the UI: Settings icon ➡ Storage ➡ New ➡ Vendor:device ➡ Hitachi HNAS
    HDS HNAS Storage management

  • Assure you have enough of disk space on the filesystem where is STOR2RRD installed
    Roughly you might count 2 - 30 GB per a storage (it depends on number of volumes, 30GB for 5000 volumes)
    $ df -g /home   # AIX
    $ df -h /home   # Linux
    
  • Test storage connectivity, either from the UI or from the command line:
    $ cd /home/stor2rrd/stor2rrd
    $ ./bin/config_check.sh 
      =========================
      STORAGE: HNAS-alias02: HNAS
      =========================
      UDP connection to "192.168.177.7" on port "161" is ok
    
      san_verify.pl:
      snmpwalk -v 2c -c public 192.168.1.1 .1.3.6.1.4.1.11096.6.1.1 SNMP version : 2c (default) Port : 161 (default) Timeout : 5 seconds Community : public Storage name : hnas01.example.com STATE : CONNECTED! 
    
      connection ok
    
  • Schedule to run storage agent from stor2rrd crontab (lpar2rrd on Virtual Appliance, it might already exist there)
    $ crontab -l | grep load_hnasperf.sh
    $
    
    Add if it does not exist as above
    $ crontab -e
    
    # Hitachi HNAS
    0,5,10,15,20,25,30,35,40,45,50,55 * * * * /home/stor2rrd/stor2rrd/load_hnasperf.sh > /home/stor2rrd/stor2rrd/load_hnasperf.out 2>&1
    
    Assure there is already an entry with the UI creation running once an hour in crontab
    $ crontab -e
    
    # STOR2RRD UI (just ONE entry of load.sh must be there)
    5 * * * * /home/stor2rrd/stor2rrd/load.sh > /home/stor2rrd/stor2rrd/load.out 2>&1
    
  • Let run the storage agent for 15 - 20 minutes to get data, then:
    $ cd /home/stor2rrd/stor2rrd
    $ ./load.sh
    
  • Go to the web UI: http://<your web server>/stor2rrd/
    Use Ctrl-F5 to refresh the web browser cache.
In case of usage of Virtual Appliance
  • Use local account lpar2rrd for hosting of STOR2RRD on the virtual appliance
  • Use /home/stor2rrd/stor2rrd as the product home
STOR2RRD uses Hitachi HCP REST API interface natively provided by the storage to get configuration and performance data.
Event logs and health status is obtained via SNMP.

Storage connectivity

  • Allow access from the STOR2RRD host to the Hitachi HCP storage on port 443 (TCP), 161 (UDP) and 9090 (API)
    $ perl /home/stor2rrd/stor2rrd/bin/conntest.pl 192.168.1.1 443
      Connection to "192.168.1.1" on port "443" is ok
    $ perl /home/stor2rrd/stor2rrd/bin/conntest_udp.pl 192.168.1.1 161
      Connection to "192.168.1.1" on port "161" is ok
    $ perl /home/stor2rrd/stor2rrd/bin/conntest_udp.pl 192.168.1.1 9090
      Connection to "192.168.1.1" on port "9090" is ok
    
  • Storage users
    1. Create global user (preferably stor2rrd) for access to the storage with Monitor role.
    Storage user management global
    2. Create additional user in each monitored Tenant with Monitor role.
    Storage user management tenant


  • Allow Management API access from STOR2RRD hosts/network/globaly for storage and each tenant
    Storage user management access

  • Enable SNMP:
    Storage user management SNMP

STOR2RRD storage configuration

  • Add storage into configuration from the UI:
    Settings icon ➡ Storage ➡ New ➡ Vendor:device ➡ Hitachi : HCP
    Hitachi HCP Storage management

  • Assure you have enough of disk space on the filesystem where is STOR2RRD installed
    Roughly you might count 2 - 30 GB per a storage (it depends on number of volumes, 30GB for 5000 volumes)
    $ df -g /home   # AIX
    $ df -h /home   # Linux
    
  • Schedule to run storage agent from stor2rrd crontab (lpar2rrd on Virtual Appliance, it might already exist there)
    $ crontab -l | grep load_hcpperf.sh
    $
    
    Add if it does not exist as above
    $ crontab -e
    
    # Hitachi HCP
    0,5,10,15,20,25,30,35,40,45,50,55 * * * * /home/stor2rrd/stor2rrd/load_hcpperf.sh > /home/stor2rrd/stor2rrd/load_hcpperf.out 2>&1
    
    Assure there is already an entry with the UI creation running once an hour in crontab
    $ crontab -e
    
    # STOR2RRD UI (just ONE entry of load.sh must be there)
    5 * * * * /home/stor2rrd/stor2rrd/load.sh > /home/stor2rrd/stor2rrd/load.out 2>&1
    
  • Let run the storage agent for 15 - 20 minutes to get data, then:
    $ cd /home/stor2rrd/stor2rrd
    $ ./load.sh
    
  • Go to the web UI: http://<your web server>/stor2rrd/
    Use Ctrl-F5 to refresh the web browser cache.