Backup & recovery
data and backup recovery solution
This seems to be a simple one as we all somehow understand backup of data and recovery of data however crude we may be doing it.
In earlier days we had a backup software installed on one server hardware and was called the backup server. The backup software had agents which were installed on different types of servers i.e. database servers, application servers, web servers etc. and information from these machines was backed up by the backup server. Now there used to be a tape drive either on the backup server or attached externally onto the backup server which used to store all the data which was being backed up by the backup server.
Then came the situation wherein the size of tapes was less as compared to the amount of data being backed up..also the tapes that were stored away with backup data – failed most of the times when data was being restored—may be due to fungus on the tapes or tapes being damaged due to non-usage etc… so essentially 80% of backups were successful and 20% of restores were successful…
Now not only server data was important – but then a situation came wherein user data also has to be backed up onto some central place… atleast for the senior management team even if not for the basic users in a company….
In IT companies documents prepared by content writers, data created by grap[hic team, applications designed or project data underway all has to be backed up.
15 years back 1GB HDD meant a huge thing.. today we carry 32GB SD chip in our mobile phones…
With growing data demand for high capacity backup also has grown … now with compliance coming into the picture – these data needs to be stored for 5 years or 7 years or may be more depending upon the company policies…
Now if a backup of a database required 10 tapes to be inserted into the tape drive one by one – there had to be one IT resource who had to stay back to insert a fresh tape onto the drive – everytime one tape finishes.. then came tape libraries which automated this process. Its like you keep a stack of 100 tapes – the robotic arm within the library picks up the required tape and feeds it into the drive….this does save time but….
" How badly you need backup is directly proportional to how badly you would suffer if you lost your data. "
Today’s backup challenges or demands
Need to backup TB’s of data within a short window period… nobody is willing to give a backup windows of 8hrs or 16 hrs etc.
Need to retain some of the data for years together.. For example a pathology lab or diagnostic clinic needs to retain the image of x-ray etc.. for at least 7 to 10 years… also more importantly – they need to ensure that this data is not tampered with – or the original image is not changed or modified.
Now with hundreds of images and other data to be backed up and stored – one would require huge amount of space – meaning tapes / LTO or huge HDD space where data can be kept.
Now with KB’s and MB’s of data – its easy to use HDD’s etc.. to store data for a long period… but how to store TB’s of data and retain the same for a long period and achieve the same within short window…
In comes Deduplication of data either at source or at the target level – how it works is – the data for example 1MB file is broken into multiple pieces… and unique pieces are backed up.. and for pieces that are found to be common only a pointer is stored – this reduces the backup space requirement to a great extent… for example you can store upto 500GB of MS word, excel, powerpoint documents within 10GB.
Now this dedupe can be done at the source level – that is from where data is being backed up or at the target—I.e. the place where data is stored.
Many companies have brought in solutions like this…. TECHRIDGE deals in EMC AVAMAR (source based backup solution), DATADOMAIN (Target based backup solution), EMC Networker – Backup Server software for complex backup environments and tape based backup requirements..
Because avamar and Datadomain store data in dedupe form creating a secondary backup copy at a distant geograqphic location is very easy as the amount of data transferred is less and the bandwidth requirement is also less.
Avamar is suitable for server backups for any database/ application including exchange, sap, oracle, db2 etc… and also for desktops or laptops.. imagine a desktop connected over a broadband connection having 10GB of data to be backed up being easily done through avamar.
Datadomain is normally used for server based backup wherein dedupe at source might use server resources and there is high speed lan available….so data is sent straight onto dd which dedupes it..avamar and DD can also be integrated together.
Datadomain also acts as a file server for storing data. Or can be configured as virtual tape library.