VMware Cloud Community
trogne
Enthusiast
Enthusiast

vsphere space issue, still growing

I don't understand why my vsphere is growing and growing.

Without any snapshots, the VM folder is over 32 gb.  It was once only 15 gb.

If you look at the "df -h" below, you'll see I'm not using that much.  Even if I substract the "/dev/sda3/" 5 gb, I'm stil at 27 gb.

So what can explains the difference between 32 gb and 15 gb ?

Here's my "df -h" :

Filesystem                                Size  Used Avail Use% Mounted on

devtmpfs                                  4.9G     0  4.9G   0% /dev

tmpfs                                     4.9G  8.0K  4.9G   1% /dev/shm

tmpfs                                     4.9G  708K  4.9G   1% /run

tmpfs                                     4.9G     0  4.9G   0% /sys/fs/cgroup

/dev/sda3                                  11G  5.0G  5.1G  50% /

tmpfs                                     4.9G   11M  4.9G   1% /tmp

/dev/mapper/netdump_vg-netdump            985M  1.3M  932M   1% /storage/netdump

/dev/mapper/imagebuilder_vg-imagebuilder  9.8G   23M  9.2G   1% /storage/imagebuilder

/dev/mapper/db_vg-db                      9.8G  108M  9.2G   2% /storage/db

/dev/mapper/core_vg-core                   25G  365M   23G   2% /storage/core

/dev/sda1                                 120M   28M   87M  24% /boot

/dev/mapper/autodeploy_vg-autodeploy      9.8G   42M  9.2G   1% /storage/autodeploy

/dev/mapper/log_vg-log                    9.8G  279M  9.0G   3% /storage/log

/dev/mapper/dblog_vg-dblog                 15G  1.9G   12G  14% /storage/dblog

/dev/mapper/seat_vg-seat                  9.8G   65M  9.2G   1% /storage/seat

/dev/mapper/updatemgr_vg-updatemgr         99G   64M   94G   1% /storage/updatemgr

Reply
0 Kudos
10 Replies
AaronRose
Enthusiast
Enthusiast

Have you considered logs?

Reply
0 Kudos
trogne
Enthusiast
Enthusiast

You mean /storage/log ?

It's only 100 mb big.

Reply
0 Kudos
trogne
Enthusiast
Enthusiast

And there was only 1.1 gb in /var/log.

So it's not that too,

Reply
0 Kudos
jhague
VMware Employee
VMware Employee

Can you provide a screenshot of the VM folder so we can see what files are using the space, e.g. is it the vmdk files or something else.

Is the VM thin or thick provisioned? If thin it will grow over time as you make changes to your VM, install patches etc. The disk will not automatically reclaim space, more info here Reclaim disk space from thin provisioned VMDK f... | HDS Community

Has the memory been increased? This would cause the swap file to increase also.

John Hague http://linkedin.com/in/john-hague | twitter @jhague10 VCIX-DCV | VCP-DCV 3/4/5/6 | VCP6-NV | VCP7-CMA | VCAP7-CMA Design
Reply
0 Kudos
trogne
Enthusiast
Enthusiast

The link shows commands for ESXi hosts.

My problematic VM is the vcsa VM, not the ESXi VMs.

In Workstations, for all the 12 disks it says "Disk space is not preallocated to this disk". That's thin provisionned.

How to reclaim disk space from the vcsa VM ?   (my vcsa is NOT inside an ESX host, it's a separate VM)

So I cannot just ssh into the vcsa and do some commands, since "df -h" shows I'm using almost anything.

I have to clean the vmdk that's in my host Windows machine, but how ?!?

Screenshot  :

Reply
0 Kudos
jhague
VMware Employee
VMware Employee

I don't see why it would grow in that manner - other than the snapshot (screenshot shows snapshot file and delta disks ...0001.vmdk but doesn't show sizes). Can it be deleted first of all?

Most space reclamation methods use a tool to zero out unused blocks and then storage vMotion to commit it and shrink the disks but that's not something you can do with workstation unfortunately. All I can think is backing up the DB and redeploying the appliance as a workaround.

John Hague http://linkedin.com/in/john-hague | twitter @jhague10 VCIX-DCV | VCP-DCV 3/4/5/6 | VCP6-NV | VCP7-CMA | VCAP7-CMA Design
Reply
0 Kudos
trogne
Enthusiast
Enthusiast

Ok.

My snapshot was already deleted. What you see is from a brand new snapshot.

Why "storage vMotion to commit it" ?  Commit what ?

How can I backup the db and import it in the new vcsa deploy ?

Reply
0 Kudos
SureshKumarMuth
Commander
Commander

could you please post the output of du -h /

above command will give you the details of the files/folder which occupies more space. will help in further isolation.

Regards,
Suresh
https://vconnectit.wordpress.com/
Reply
0 Kudos
trogne
Enthusiast
Enthusiast

With "df -h", the size of "/" (dev/sda3) is 5 gb out of 11 gb.

But my vmdk for this disk, with no snapshot, is 8.71 gb out of 12 gb.

There is 3.71 gb lost just there.

Later I'll get back with the output of du -h /

Reply
0 Kudos
trogne
Enthusiast
Enthusiast

Here it is :

du -h / --max-depth=1

0       /dev

220M    /etc

244K    /home

100K    /tmp

2.6M    /tftpboot

712K    /run

12K     /mnt

473M    /opt

0       /sys

1.1M    /vasecurity

26M     /boot

56K     /root

3.2G    /usr

0       /proc

3.9G    /storage

16K     /lost+found

1.2G    /var

8.9G    /

So, the total machine is only 8.9 GB.  But the sum of all my vmdks is 32.9 GB.

Which is 24 GB lost. No way at all to get back all this space ?

Reply
0 Kudos