3 commands for easy Nimble

OK, just a small thing I’d like to share:

When checking an iSCSI Nimble setup, 3 tiny commands can make your life a lot easier. In VMware ESXi 5.lul, SSH to a host and get your storage straight:

1) If not already done so: change the default Path Selection Policy to Round Robin for new disks:

esxcli storage nmp satp set –default-psp=VMW_PSP_RR –satp=VMW_SATP_ALUA

2) Set all your already connected Nimble disks, and only Nimble disks, to Path Selection Policy RoundRobin (VMW_PSP_RR) Handy in a mixed storage provider environment!

for i in `esxcli storage nmp device list | awk ‘/Nimble iSCSI Disk/{print $7}’ | sed -e ‘s/(//’ -e ‘s/)//’`; do esxcli storage nmp device set -d $i –psp=VMW_PSP_RR;done

3) Set the IOPS to “1”in the Path Selection Policy VMW_PSP_RR for each connected Nimble disk as a new recommendation from NimbleStorage

for i in `esxcli storage nmp device list | awk ‘/Nimble iSCSI Disk/{print $7}’ | sed -e ‘s/(//’ -e ‘s/)//’`; do esxcli storage nmp psp roundrobin deviceconfig set –type “iops” –iops=1 –device=$i; done

More info for ESXCLI command options at http://pubs.vmware.com/vsphere-50/index.jsp?topic=%2Fcom.vmware.vcli.examples.doc_50%2Fcli_manage_storage.6.5.html
Discussion from Cormac Hogan at http://cormachogan.com/2013/07/08/automating-the-iops-setting-in-the-round-robin-psp/


I asked Nimble how they got to the IOPS=1 part, and they said:

“This is as the result of extensive internal testing. We have nothing available for customers, this was changed due to an issue we identified when migrating host profiles and is also set automatically by our NCM plugin.”

Dell iDrac FW 1.50.50 picky on time formats

As I recently discovered, a small “oops” on behalf of Dell:

New Dell PowerEdge servers are being shipped with iDrac firmware version 1.50.50. This specific firmware version has the unfortunate side effect that when you open the iDrac webpage (default IP address in IE11 with incorrect time and date settings, the login page is displayed without any Java elements. You basically can not log in.

In any browser, on any OS, the login page is displayed without any Java elements. Regional Date/time formats are NOT “English (United States)”. Login elements are not available, extra information is not available, nothing can be done in the page. Dell logo, some login boxes and some plaintext is visible.

Date/time formats are NOT “English (United States)” which causes Java elements not to run.

Go to Control panel – Region – Tab Formats. Change the date/time format to “English (United States)”. Reload the iDrac web pagem and all the elements in the login screen do work again. MIND YOU: This can cause legay or poorly written applications to bounce, break or just stop working. Think about the impact before you change it.

Downgrade the 1.50.50 verion to an earlier one, and wait for Dell to fix it.

Dell has confirmed this is a bug in their brand new fw 1.50.50 version. They are working on a new version in which this will be fixed. be patient, my young OutOfBand administrator ..

-+* Update *+-
Dell has released 1.51.51 which fixes this bug, amongst some others. See http://www.dell.com/support/drivers/us/en/19/driverdetails?driverid=JGG1X for more details.

Shameless Call for Hardware


In the past 6 months, I have had numerous discussions about the use of old hardware, or plainly put “Iron”,  in test / playground environments. Whilst virtualisation guys emphasise the need to virtualise everything, I am still an advocate of putting old and discarded hardware to use. (All hardware? Of course not. Read on.) For several reasons:

1) Even when virtualising the hell out of everything, you still need Iron to run it on. VMware ESXi or Workstation? Storage, CPU and RAM needed, thank you. Storage product simulators or other virtual appliances? Same here.

2) Iron, often of any age, still can provide functionality that extends beyond any virtual product available to mere mortals. An Equallogic Software Simulator is extremely hard to find. or, try to virtualise a complete FC fabric. Feel how an LC connector slides into an SFP. And so on.

3) Very often people like to touch things. They need to feel the hardware to get the idea behind the product, or technique. And yes, most pocket devices, phones or tablets have more computing power than an old NetApp FAS270, but the Iron still provides the possibility to design, build,  troubleshoot and perform destructive testing on. Something that software appliances mostly do not offer.


My current employer now gives me the opportunity to build two 19″ racks with iron, that can be used for a number of reasons:

– Create a reasonably representative environment with Storage, Gigabit Ethernet connectivity, VT-d capable servers and an FC fabric to use for testing, troubleshooting and replicating customer problems

– Create a software testing ground for elaborate and extensive testing of for example Group Policies, Citrix settings and Exchange optimalisation in a Known Environment

– Perform research for new products. Create installation manuals, best practices and Sales documents for products that can be added to our own Product Catalog.

– Offer an opportunity for technical young guns to expand their technical knowledge, and get used to working with project skills such as Time Management, Documenting and Reporting to stakeholders.

But, why all this? Well, let me refer back to the subject of this post. I have experienced in the last year that lots of hardware gets discarded every day. Perfectly working order, licenses available, functionality present .. and yet, it gets thrown in /dev/null. Well ..


I am very much looking for all sorts of old hardware. I can make it of use for our internal education department, I can create environments that can (up to a reasonable level) represent corporate setups, etc etc. So, OK, here it is .. please contact me when you have some stuff collecting dust on the shelf, and you think that it still has some capabilities.


I cannot collect and dust up every single time something is discarded. Even I have some (what I think reasonable) demands:

– For Servers: rackmount or blade center servers that are capable of running VMware ESXi 4.1 or higher
– For Storage:  NetApp (FAS3XX0 and higher), Compellent or EqualLogic hardware
– For switching: Managed Gigabit
– For Fiber Channel:  2gbit or higher is usable
– For tape products: LTO3 or higher standalone or libraries, FC or iSCSI
– Disks: 2.5″ or 3.5″ SAS or FC
– Anything else that can remotely possibly be of any use


 I’d really appreciate it if I can collect some things to get my hardware test / educational environment up and running with stuff that is in one way or another representative for  corporate environments.


Of course. But, otherwise I could never ask any of this to you. Thanks will be sent in return in the form of cake or beer.

Thanks in advance, cool stuff!

Cisco MDS 9120 Password recovery

In spare time, I often try to pry open old hardware that we received from customers. Most of the aren’t that hard, but some are more challenging. For example a Cisco MDS 9120 Multilayer FC switch. This one is totally different from the McData Sphereon 4500’s or EMC DS-24M2 and Brocades that I normally work with in my lab. Normal recovery procedures with the configuration register from Cisco (http://www.cisco.com/en/US/docs/routers/access/2600/hardware/installation/guide/2600appC.html) won’t work either. OK, let’s google this, then.

http://www.scribd.com/doc/11415473/Cisco-MDS-9000-Configuration-Guide#outer_page_346 seems a nice hit. Just hit CTRL-j in the boot sequence and the switch(boot)# prompt appears. Right. But, after trying 5 times to get into the boot prompt, still no results. Back to the documentation. Read carefully: It’s not CTRL-j, but CTRL-]. I will try this tonight.

I will read the documentation better.
I will read the documentation better.
I will read the documentation better.
I will read the documentation better.

Dell Storage Forum

Hurray! I’ll be attending the Dell Storage Forum at Paris! Really looking forward to this event. Nice Pre-conference courses, great technical and advanced sessions, lots of people to meet and more.

Again NetApp: Metrocluster Workshop

One of the advantages of being a Silver Partner, is that you can attend free courses, workshops or other (pre)sales sessions organised by AVnet. Last friday I attended the NetApp Metrocluster workhop at their Almere office. First part was a bit of cabling, which scared some visitors off. They were expecting the full mont right away. And allthough the presenter tried to explain that that was up next, the guy disappeared right after the first coffee break. Pity for him, he missed the opportunity to get some info about metroclusters from one of the best metrocluster guys NetApp Europe has to offer. Session had more technical oomph than I expected, so that’s a good thing. FC protocol dripped through as well, so I had a nice day! Unfrtunately, it ended rather quickly, 2pm at a fridayafternoon! Ah well, beer at the main office then ..

NetApp Silver Partner Course – Time Well Spent?

I recently have enrolled the company I work for in the NetApp Silver Partner Program. And, as one of the prerequisites put by the Distri, at least one of the employees of a new Silver Partner company has to follow a NetApp Silver Partner Course organised by the distri. Me and a colleague from work were attending this course, with a total of 5 participants. The teacher should have checked the technical level he was up against, us 2 were way more developed in our NetApp fu than the others. In short: you can learn a lot of the product if you’ve never or nearly never heard of it. WHen you are experienced with it, it’s not recommended. Or a waste of your time. On the other hand, I can understand that a distri / producer wants to have a decent and equal level of knowledge from alll their new partners. It’s a tick on their checklist. So, that’s how I see it. A waste of time, but necessary.

Storage Expo 2011

Looking forward to Thursday 3rd of November! The Storage Expo is then taking place at Jaarbeurshallen Utrecht, together with Tooling Event and InfoSecurity. Last year it got me a job at a very exciting ICT-services company. This year I want to meet the guys from NetApp and Veeam, and see what other people can share at the various seminars and talks. See more at http://www.storage-expo.nl. And, best of all, it’s free! Keep an eye on the badges, you might see me!

Migrating BackupExec 2010 R3 to new server

OK, facing one very interesting project: migrating a complete BackupExec installation with all data, jobs and history to a new server. I will follow This Symantec Article , see where it will bring me! More news in a couple of days ..

ADDED OCT. 13th:

OK, I succeeded! here are the steps necessary to migrate stuff:

– Make sure that source and destiantion BackupExec versions are the same. It’s essential for Database schema (or mismatches, if versions are not the same)
– Run LiveUpdate on source and destination until no more updates are available. Then run it again, just to be sure. LiveUpdate can have its weirdnesses
– Make sure no jobs are running at the time of the data migration. if necessary, put alll jobs on hold


OK, there we go! Source is prepared, destination is ready, migration time!

1) Write down all license information (Tools – Install options and license keys on this media server)
2) Write down the Backup-to-Disk location (In my case an iSCSI LUN on a small SAN)
3) Stop all Symantec BackupExec 2010 services. Do this through the BE2010 console, or through services.msc. Also stop the BE2010 SQL Server service!
4) All the BackupExec data is located in c:\program files\symantec\backupexec\data. copy this to a temporary location, available to the new backup server. I assume everyone is comfortable enough with share$’s. Do not move the msgq*.* files!
5) Run through the destination install of BE2010 again, checking that LiveUpdate is done etc etc.
6) Stop all BackupExec services, and the BE2010 SQL Server service.
7) Copy the saved \data files from the source to c:\program files\symantec\backupexec\data in the new BE2010 install
8) Rename the DataPartition part in the SQL Database:|
8a) Open a CMD prompt and type “Sqlcmd -E -S servername\BKUPEXEC”
8b) Type “use BEDB” and type “go”
8c) Type “SELECT partitionname FROM datapartition” and type “go”
8d) Type “UPDATE DataPartition SET PartitionName='<Dest.server name>’ WHERE PartitionID =0”
8e) Type “quit” and exit the CMD prompt
9) Start all services
10) Check if all Job Data is correct and remove old devices
11) Reconnect the iSCSI LUN in your Windows server, and recreate the Backup-To-Disk folder in BE2010.

And you should be good to go! In rare occasions, you can run into errors while selecting the new Datapartition on your SQL database. Try to run the CMD prompt as admin.

Reboot the system one more time, to check if every service starts OK. Oh, and don’t forget to update your mailserver to allow mail relay from the new server IP 😀