How to reduce coredump size in Solaris 11

How to reduce coredump size in Solaris 11.

Suppose rpool is full and we don’t have any scope for further housekeeping then we can try to reduce the size of coredump to free up some space in rpool.

If we reduce the size of coredump directly it will give us error and ask to do it forcefully but doing it forcefully may cause data loss.

Below are the steps to do reduce coredump size.

#dumpadm     >>> to get info on current dump device and savecore directory

# zfs get volsize rpool/dump       >>> To get vol size of dump device

#zfs create -V 10G rpool/dump1       >>> creating new dump volume of size 10G (suppose previous dumpsize is 20G)

# dumpadm -d /dev/zvol/dsk/rpool/dump1          >>> set the dump device to new volume

#dumpadm             >>> We can see new dump device

#zfs list |grep -i dump                >>> We can see 2 dump devices here

#zfs destroy rpool/dump                       >>> destroy old dump device

Thanks,

Kiren Jadhav

How to split big files in Unix

 How to split big files in solaris:

Sometime the log files is too large to send backend team for further analysis, in that case we break the files into multiple pieces using split command.

Whenever we split a large file with split command then split output file’s default size is 1000 lines and its default prefix would be ‘x’.

Suppose we want to send crash dump (vdump) file which is 2GB of size. We will break that file into 4 pieces of size 500MB.

In that case:

#split -b 500M vdump.24 = This will break vdump.24 file into 4 pieces of name Xa , Xb, Xc, Xd

If we feel aa ab are little bit confusing file names then better to give specific name at the time of spliting the file itself.

#split -b 500M vdump.24 vdump_11May2021_X = It will create new files of name vdump_11May2021_Xa ,  vdump_11May2021_Xb

If we want to split file into GB then we may use command something like:

#split -b 1G vdump.24 vdump_11May2021_1

or

#split -b 1024M vdump.24 vdump_11May2021_1

Regards,

Kiren Jadhav