Hi, my name is Alan Johnson and  I have been working in the data storage industry for over 35 years. My areas of interest include tech writing and investigating new areas of research such as the latest storage initiatives. Published books are available at http://www.lulu.com/spotlight/alan52/ or you can ask me for a soft copy of my “Data Storage Fundamentals”

View Alan Johnson's profile on LinkedIn
Thanks for reading,let me know if there are any pages that you would like added,  and I can consider putting them into the lengthy queue. My writings are done as a hobby and I can only add content as time permits.

If you find these tutorials useful please leave comments or send an email to me. Also please feel free to link to any of the pages in this site. I cannot always respond to technical questions as some of these were written a while ago. The intent is to share my experiences and the postings here are in no way authoritative.

Thanks for visiting!


14 thoughts on “About

  1. thank you for the articles. I’d like to see one on FC. short theoretical intro that explains addressing, fabric, use of FC switches and then a practical multipath setup that anybody can try with a couple of computers and cheap ebay fc cards. whether with or without a FC switch (brocade 300 on ebay is also cheap enough for people to buy and practice on)

    i recently bought 3 dualport qlogic fc cards, put them into my pc and 2 hp microservers and i’m finding it quite hard to even start. It is unbelievable how little in the way of literature is available for people who don’t want to study for single vendor’s certification program.


  2. The articles are pretty much useful to lot of folks who are looking for the information. I will expect more articles on storage areas.


  3. Hi.
    I have a trouble in ceph hands-on guide
    all go ok until i do de ceph deploy from monserver to osdservers.
    it give to me this error

    [monserver0][WARNIN] –2016-06-20 16:14:18– https://download.ceph.com/keys/release.asc
    [monserver0][WARNIN] Resolviendo download.ceph.com (download.ceph.com)…, 2607:f298:6050:51f3:f816:3eff:fe71:9135
    [monserver0][WARNIN] Conectando con download.ceph.com (download.ceph.com)[]:443… falló: Expiró el tiempo de conexión.
    [monserver0][WARNIN] Conectando con download.ceph.com (download.ceph.com)[2607:f298:6050:51f3:f816:3eff:fe71:9135]:443… falló: La red es inaccesible.
    [monserver0][WARNIN] command returned non-zero exit status: 4
    [monserver0][INFO ] Running command: sudo apt-key add release.asc
    [monserver0][WARNIN] gpg: no se han encontrados datos OpenPGP válidos
    [monserver0][ERROR ] RuntimeError: command returned non-zero exit status: 2
    [ceph_deploy][ERROR ] RuntimeError: Failed to execute command: apt-key add release.asc

    any suggestion?


    • It looks like you are having trouble contacting download.ceph.com. Try and ping it to see if the networking is set up correctly. You can also check with the docs at ceph.com to double check – AJ


      • i can ping i try to do this un a remote server . i’ll talk with the admin to ask if i have any trouble with port 443.
        thx for reply
        And for every great guide you’re doing.


  4. Hi Alan,
    Hope you are doing well. I couldn’t find any way to contact you so just thought of commenting here I’m looking for an opportunity in Big Data/Hadoop so just thought of having a word with you if I may stand any chances with Supermicro.

    Looking forward to hear from you. Thank You.


  5. Alan, thank you very much for your tutorial: configuring-separate-ssd-and-hdd-pools-with-ceph-mimic, it has served me a lot, I just want you to help me clarifying if it is strictly necessary to create an erasure-code-profile Eg. “ssdprofile”, or with the “default” it can also work well? And what is the function of the ecpool? regards


    • Hi Felipe, it is not necessary to create EC profiles, I did it just to show more examples. I have found though that a lot of users will not use 3 way replication with SSDs as the cost is too high so they will either use 2 way rep or ec to reduce the expense.


      • AJ, thank you very much for your response!
        But I have a doubt, what exactly is the function of an ecpool? Or can I work an ec profile without setting up an ecpool?


        • Hi Felipe, to store data it is necessary to create a pool, There are two types of pool with regards to data integrity – replicated pool and erasure coded pool. You can have both created at the same time if you want. A replicated pool makes three copies of the data (by default) across three of the OSDs, so if you have 9 disks you will only have usable capacity of three OSDS. If you create an erasure coded pool it is much more efficient, for example if you use a 7+2 code across the 9 disks you will be able to use 7/9 of the total capacity whereas with 3 way replication only provided 1/3 of the total capacity. With a 7/9 scheme you could lose two servers and still carry on. Note you will need a total of 9 servers (11 are recommended for a 7+2 erasure coded scheme (unless you modify the crushmap). The default is to replicate data on different disks on different servers so the resiliency is at host level. The crushmap can be modified to allow the resiliency to be at OSD level but this is not recommended in production environments.


Comments and suggestions for future articles welcome!

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.