EVA Max Transfer Size

Another HP EVA Post from the vaults:

Recently I saw an issue where an EVA was delivering poor performance when under load from a full backup. I used EVAPerf to investigate the issue and found that the disks were not under pressure, however the CPU was running up to 100% busy. During these spikes of high CPU usage the latency on the LUNS was inevitably very high.

 

I found the following advisory from HP, which describes performance issues which are created by the MAX IO limitations of the EVA. THE EVA was designed to deal with a max IO size of 128KB, when receiving a transfer size larger than this it has to buffer it and then break it down, putting an increased load on the system. The suggestion from HP is to limit the max IO size from the host system. The hosts connected to the EVA in question were Windows 2008 R2, the method to limit the IO size was via the HBA.

 

The hosts I was working on were using Qlogic HBA’s. To make the changes to a Qlogic HBA you need to be on a driver version of 9.1.8.28 or later and then follow these steps from the Windows command line:

1 View current MAX IO size qlfc -tsize /fc

2 Set MAX IO size to 128KB qlfc -tsize /fc /set 128

3 Set as default qlfc -tsize /fc /set default

 

Since implementing the reduced block size the performance has returned to a good level.

EVA Upgrade

This post covers the upgrade process for a EVA Upgrade.  I wrote it ages ago but never got round to publishing it.  In the spirit of blogtober I wanted to get it out there as I am sure there are still some HP EVA’s in use .

This is the steps I followed to upgrade the firmware on an EVA. You can read further on the steps using the EVA upgrade guide

Week Before

  • Check current EVA and Command View versions
  • Upgrade Command View to compatible version if relevant -10.3
  • Download firmware from software depot
  • Check for any hardware failures
  • Upgrade HBA and switch firmware – check compatibility with SPOCK
  • Check disk grouping policy is set to manual. System options, configuration, set system operation policies, manual
  • Ensure you have The EVA SSSU (HP Storage System Scripting Utility) the command line tool for managing EVA systems installed. SSSU will generally be installed on your EVA management server
  • To Use EVA SSU launch the program from the shortcut. You will then see a screen like belowSSSU for HP P6000 Command View

    Version: 10.2.0

    Build: 091012A

    Manager:Name or IP address of CommandView Server

    Username:administrator

    Password:

    Under manager enter the IP address or hostname of you EVA management server, then log on with an account with rights to the system. You can check which accounts have rights by looking at the local security group xxxx on the management server and seeing what local accounts are in there.

    Once logged in you will need to choose which system to manage. To see what systems are available first list them:

    NoSystemSelected> ls system

    Systems available on this Manager:

    EVA01

    Then select which system you want to manage.

    select system EVA01

Day Before

  • Check hardware
  • Check events – controller events
  • Backup EVA system config as follows:

First connect and logon to the system using EVA SSSU instructions as described above

To capture system config to a file:

capture configuration C:config.txt

You can use the capture validate command to verify if the contents of the file still match the current configuration of the EVA.

Capture validate <filename to check>

 Also as a best practice run ls disk_group full, ls host full, ls vdisk full before an upgrade

ls disk_group full > C:disk.txt

ls host full C:host.txt

ls vdisk C:vdisk.txt

  • Backup data on EVA

Day of Upgrade

  • Check hardware and controller events
  • Configure User-Initiated Service Mode (UISM). System options, configuration, configure User-Initiated Service Mode (UISM)
  • Disabling array passwords. Settings, management options, security options – manage storage system password access
  • Shut down hosts – if offline
  • Upgrade controller software – system options, code load system
    • Tick box confirming you have completes pre-upgrade checks
    • System runs automated system checks to see if its fit to continue
    • Code transferred to controllers
    • Upgrade begins
    • Nodes reboot
  • Verify version of XCS firmware after upgrade
  • 30 min after upgrade check IO module firmware on disk enclosure
  • Upgrade drive firmware allow 30sec-3 min per drive

A Quick Guide to EVAPerf

This week I had to take a look at some performance concerns on an EVA. I had to dust off the old EVAPerf skills and made some notes along the way which I thought I would share.

 

EVAPerf is a command line utility used to monitor the performance of an EVA. It can be used in a console mode for real time monitoring or have the output piped to a CSV file for future review.

To run EVAPerf you need to connect to the server its installed on, open up a command prompt and change the EVAPerf directory e.g. cd C:Program Files (x86)Hewlett-PackardEVA Performance Monitor. The key command to collect all performance stats and pipe them to a CSV is:

evaperf all -csv -cont <xx> -dur <yy> -ts2 > <outputfile>

all = collect all performance counters

-csv = make file CSV type

xx = the interval at which data is collected in seconds

yy= how long (the duration) the collection will run for in seconds

-ts2 makes the file follow Microsoft Time format

I was troubleshooting an issue which occurred at a particular time each day and so created a batch file with the commands below and scheduled it using Windows Task Scheduler.

FOR %%A IN (%Date:/=%) DO SET Today=%%A cd “C:Program Files (x86)Hewlett-PackardEVA Performance Monitor”evaperf all -csv -cont 60 -dur 3600 -ts2 >f:%Today%.csv

The above command will give you all performance counters, for a duration of one hour, will collect stats once a minute and pipe the results to a CSV file with the days date.

If you want to see which individual performance counters are available just type evaperf with no options and press enter, this will then give you a list of all the available commands you can use to monitor the EVA. Some of the most useful ones are:

evaperf pd = physical disks stats 

evaperf vdg = virtual disk group stats 

evaperf vd = virtual disk stats  

evaperf hps = host port stats 

evaperf cs = controller stats

To monitor in real time just enter any of these commands directly into the CLI.

 

To keep up to date with more news follow 3ParDude on Twitter