Hello friends, please let me Go take a hand.
I have a client that has a VMware ESXi 4.0 running on an IBM server, there a few daysago the server stopped working for lack of hard drive space, so IT techniciansbecause the company made several unnecessary snapshots. I went there, delete thesnapshots and even VMDKs and other invalid files, released 30GB, but now the serverreturned to lose ground and now has only 12MB of free disk space. How can thishappen? What makes a server loses much disk space so quickly?
Thus, the virtual machines are stopped, the server crashed Vmware all what I do, how to free space without losing data?
Thanks
edit by continuum
all attachements were removed by OP so if you read this post keep in mind that the steps used to resolve the issue no longer make sense.
If you run into similar issues you can not learn from this post
Ulli
looks good - everything seems to work as planned
if you want - attach the vmx after you resolved this issue so that we make sure it does not happen again
Hello Andrew, in 20 days or so, the client is another server here on site, this server is a Dell Power Edge T610, new hard drive with 900GB of RAID-5 and 16GB of RAM. In itwill install the latest version of ESXi and move all the virtual machines to this server.The current IBM I'm doing the procedure for consolidation of VMDKs will be formatted, I'll install Windows Server 2008 with Veeam Backup & Replicator for backing up virtual machines and files contained in VMDKs of all virtual machines, this clear with a Storcenter IX-200 Iomega 4TB iSCSI to store all these backups.
That is, in 20 days the things here in this account will be a little more professional,since many errors were caused by inexperienced technicians, for example, if they wereto install an HP printer driver on a virtual server, before they did a snapshot, virtual machines in this game had at least five snapshots accumulated.
Well, now I'm here trying to solve the problem, but I think the solution given by theContinuum will work, the process is 55%, though.
Sorry about the lack of experience, but what do you mean by attach the vmx?
Friends, for security reasons I removed all the attachments in my posts. Over the next ifyou have the need to send files, I'll do this with private messages, okay?
Thank you for understanding.
Hello,
All you need to do is to edit the VM's properties and attach the new virtual disk as the second HDD to your VM. Before editing the properties, open the datastore browser and download the VM's RDS-FS.vmx file for backup purposes! Once everything is ok with the new virtual disk you may delete the old files.
Caution: Make absolutely sure you only delete the following files! Do NOT delete RDS-FS.vmdk, RDS-FS-flat.vmdk and all the other files!!
The "_1" makes the difference!
RDS-FS_1-000001.vmdk
RDS-FS_1-000002.vmdk
RDS-FS_1-000003.vmdk
RDS-FS_1-000004.vmdk
RDS-FS_1.vmdk
RDS-FS_1-000001-delta.vmdk
RDS-FS_1-000002-delta.vmdk
RDS-FS_1-000003-delta.vmdk
RDS-FS_1-000004-delta.vmdk
RDS-FS_1-flat.vmdk
Once everything looks ok power off the VM and copy (don't move) the virtual disk back to the original location.
Don't forget to modify the VM's properties again after copying the virtual disk.
André
BTW: After cleaning up this VM you should have enough free disk space to cleanup the WSUS VM.
Friend, before answering this last post, I had already started the process, did thefollowing.
a) I removed the disk from the VM without deleting the files from the datastore.
b) I created a new virtual disk using the new file fs_data.vmdk.
c) I started the virtual machine, did tests and saw that all data is it, all right, no losses.
d) Delete all VMDKs, not the RDS-FS.vmdk which is the virtual machine's operating system, just delete the virtual disk VMDKs problematic.
e) And right now I'm moving the new file to the original datastore, I used the copy option, it's really a risk and you are right, but now I went and monitoring, is still 9%,because the disk has 150GB Storage and a FreeNAS is being moved to the hard diskfrom ESXi host, the network here is 100Mbits only then is wait.
Well, you hope everything goes well.
Thanks.
Looks like this is going to be another long day for you
André
Again? Friend, began consolidating at 9:00 AM, ending at 5:15 PM.
Right now moving the file is 26%, well, I think the rhythm that ends by 8:30 PM, I guess.
Until then I'll be studying here.
If the network was gigabit, it would be much faster, but patience and more patience,are now the final moments
Ivanildo Teixeira Galvão wrote:
Friends, for security reasons I removed all the attachments in my posts.
Too late - I already have the keys for your Porsche and emptied all your bank accounts
Serious again ... there is really nothing that could be exploited in such data.The problem with sending such data via emails is that you will only get help from one person - but I am really not the only one who can help in such a case.
Anyway - I respecty your wish to maintain your privacy.Back to work ....
the steps you outlined are ok - proceed like thatI edited your first post to tell other users with similar issues that they can no longer learn anything from this post as the attachements were removed
Ulli
I will edit the images to omit some information and put all again, ok?no - don't waste any time for that - its ok
Question, after all is over and virtual machine running. What will I need to do to ensure that this new 150GB vmdk now has not re-grow wild?
Depends - if you use Veeam or similar you have to make sure that no disks are set to persistent mode.If you use no automatic backuptools you can limit the max number of snapshots that you want to allow.
You can also set a limit by setting a minimum of free space that must be available after taking a snapshot.It really depends on your use-case and wether you plan to use automatic backup-tools
Read about useful parameters on my site
http://faq.sanbarrow.com/index.php?solution_id=1060
or tell us what you want to allow and what not - and I can suggest settings
you can for example set
snapshot.numSnapshots = "1"
to avoid that your colleagues/ customers take new snapshots too carelessly