There are are two backup jobs out of 19 jobs that are running extremely slow if i run them on different schedule.
Backup Tool:backup exec 2012
Tape Library :quantum scalar(LTO5)
I was unable to run those jobs on last weekend which was there scheduled run.Now i trie to run those jobs on this weekend but the spped is extremely slow.I checked the history and found that this is common problem for these backups.
I chceked with storae team they said there is no issue from there side.My drive is directly connected via FC.
Kindly advice what could be the probable reason and where i could change any settings to improve thorughput.
Are you comparing NDMP jobs with non-NMDP jobs when you say that two jobs are slow (out of 19)?
If yes is your drive connected to the Backup Exec server or the NDMP device (or both) via FC? If only connected to the BE server than your data is being transferred over the LAN which is not the optimal way to use the NDMP option.
If the drive is locally connected to the NDMP device then as Backup Exec sends an NDMP command to run the backup and then teh NDMP devcie handles teh command itself, other than monitoring the status Backup Exec is not involved with the data transfer between NDMP Disks storage and tape storage and performance issues are then potentially between the hardware/OS of the NDMP device and the tape system. Of course if all 19 selections for backup come from the NDMP Disk storage then it may be the nature of the data held in the different selection areas (or the volumes holding the data) that is making the difference and you may need to start comparing in those areas.
All the 19 jobs are for NDMP backups using the same tape library.
The data i am backing up is the vaulted data.If i need to check the volumes which are holding data.How can i check the performance for that.As i am assuming that the tape drive is not having any issues.when i run the backup just after cancelling that job i can see that the data is coming from same filer but different vault location and its clocking amazing speed.
as i highlighted i already raised a case with storage team to check on this there reply was straigh forward that the io ops are stable.
Not sure what you mean by vaulted data, however if the issue is clearly related to the data or the data location it is unlikely to be something any settings within Backup Exec will resolve
It may even just be the nature of the data itself
i.e. 1,000 files of 10MB each (so 10GB of data) will usually backup slower than 10 files of 1,000 MB each (which is also 10GB)
I agree to these facts but when those backups are run on specific dates as per schedule then they run with good spped but when they are run on random dates apart from there schdeuled date.It goes for a toes .
Its taking the backup from seconday location not the prod.