December 29, 2018 at 8:30 am
Hi All,
I honestly think this is a great platform to learn and I am new to database administration( since 3 yrs do not count as much) with good experience in testing and team management. Now, I like SQL server and google does help sometimes to resolve few alerts :). Of course sometimes it is here where I get the help. Now, I am seeking advice here from experienced professionals.
I study a lot, learning different technologies and certifications like AWS, Azure, ITIL etc. Now, my concern and worry point is the on call support and how to prepare myself for downevents, what will I do when db's hit corruption, How to properly migrate databases ( I do have script ready for orphan users :). I never built a cluster but would like to build a 2 node cluster. But since I do not have test environment what are my options, can I build one at my home or lab? How did you manage?
Now do you prepare yourself for these time sensitive events. It scares me sometimes and wonder is DBA actually for me or go back to testing? When I was doing testing though I always had a feeling of not working or building stuff and that made me play around with databases. Please advise and thank you!
December 29, 2018 at 8:43 am
sizal0234 - Saturday, December 29, 2018 8:30 AMHi All,I honestly think this is a great platform to learn and I am new to database administration( since 3 yrs do not count as much) with good experience in testing and team management. Now, I like SQL server and google does help sometimes to resolve few alerts :). Of course sometimes it is here where I get the help. Now, I am seeking advice here from experienced professionals.
I study a lot, learning different technologies and certifications like AWS, Azure, ITIL etc. Now, my concern and worry point is the on call support and how to prepare myself for downevents, what will I do when db's hit corruption, How to properly migrate databases ( I do have script ready for orphan users :). I never built a cluster but would like to build a 2 node cluster. But since I do not have test environment what are my options, can I build one at my home or lab? How did you manage?
Now do you prepare yourself for these time sensitive events. It scares me sometimes and wonder is DBA actually for me or go back to testing? When I was doing testing though I always had a feeling of not working or building stuff and that made me play around with databases. Please advise and thank you!
IMHO, being a good DBA means hoping for the best whilst planning for the worst, no backups are valid unless tested, no hardware is failure proof, all users are idiots, developers are the worst idiots, management does not have a clue and nothing is fool proof as there will always be better developers/users.
😎
Automate everything that can be automated, i.e. a restore should be pointing a script to the backup folder, no manual work. This way, you can mitigate and offset the time sensitive factor and minimize potential errors in a stressful situation. A good example from few years back, a multi tenant SaaS system went down because of a developer's cleverness, which filled the tempdb. It took 5 sec. to recover the functionality as I had already anticipated this situation and had stored procs adding additional files to the tempdb in place😉 Non of the customers ever complained!
December 29, 2018 at 10:57 am
Thank you and appreciate your advice. Time to learn more.
My Weekends for past couple of months have been to learn and learn more.
It is amazing and stressful sometimes.
December 29, 2018 at 10:58 am
In addition to automation do you recommend any books or strategy which helps in moving towards senior dba?
December 30, 2018 at 10:26 pm
sizal0234 - Saturday, December 29, 2018 10:58 AMIn addition to automation do you recommend any books or strategy which helps in moving towards senior dba?
If you have a minimal scripting knowledge and basic commands to combine etc. You can learn.
As long as you have a system and internet, you can learn and test everything in your home. Install multiple instance and have a test upgrade, migration , clustering & AG.
For automation take one server as a centralized server and create a linked server and to pull the data to CMS like, disk space of all serves in onle palce, there you go 🙂
Muthukkumaran Kaliyamoorthy
https://www.sqlserverblogforum.com/
January 1, 2019 at 1:17 am
Thank you for the advice.
January 1, 2019 at 6:41 am
inshah2297 - Tuesday, January 1, 2019 5:14 AMBeing a decent DBA implies seeking after the best while getting ready for the most exceedingly terrible, no reinforcements are substantial except if tried, no equipment is disappointment evidence, all clients are imbeciles, designers are the most noticeably bad morons, the board does not understand and nothing is secure as there will dependably be better engineers/clients.
Robotize everything that can be mechanized, i.e. a reestablish ought to indicate a content the reinforcement envelope, no manual work. Along these lines, you can moderate and balance the time touchy factor and limit potential mistakes in an upsetting circumstance. A genuine model from couple of years back, a multi occupant SaaS framework went down due to an engineer's astuteness, which filled the tempdb. It took 5 sec. to recoup the usefulness as I had officially foreseen this circumstance and had put away procs adding extra documents to the tempdb in placeWink Non of the clients at any point grumbled!
Is that a Google translation of my previous post?
😎
January 1, 2019 at 7:17 am
When I started out in data, I was lucky enough to hire a SQL Server DBA veteran of like 10+ years experience. He was what I consider a legit DBA that focused more on the setup, management, and recovery of data critical systems that if they go down, lives could be at stake.
He would always tell me that in the time a inexperienced DBA was Googling the answer, veterans like him would be executing a fix for the problem. He would already be running the race, while you are still stuck at the starting line trying to figure out how to run. The time lost looking up the answer could cost the business thousands if not millions of dollars or worse.
Thus, if time sensitive tasks scare you, then it should. The only way to get better is to practice. This means, if you have to do a point-in-time recovery, then you better learn how to execute it without having to look it up on Google when it happens. You can practice on a test environment and mimic having to do this regularly to ensure you can both do it from memory as well automation. The same is true for any other disaster scenario. You should always be doing regular testing such as disaster recovery, penetration testing, etc with you and the team. Don't rely on just scripts, but also doing it without scripts. Even without the fancy tools with GUI such as executing SQL queries from command line remotely.
When it comes to learning new things, then one good way is using Azure or AWS cloud computing. Setting up a cluster on the free tier is a common practice. Even shelling out the cash for the lower end CPU tiers to make a 10-node cluster that maybe costs you 20 USD is not uncommon. I did that a lot to learn how to make large clusters of Hadoop and so forth. The money spent there is a investment in you and your learning.
Viewing 9 posts - 1 through 8 (of 8 total)
You must be logged in to reply to this topic. Login to reply