|
Vulture Culture posted:I'm nowhere near an expert anymore, but failing anything else, I'd pass the drive through to a VM and SCP the data over to one of the connected hosts via the Tech Support Mode SSH server. Unless you've got huge sparse files, I don't think there's any particular reason you would need to use VMFS on your removable media -- just download the disks and VMX files. Thats the rub, I can passthrough an NTFS formatted usb drive to a new VM but now what? The usb drive isn't bootable, the new VM is blank and because of usbarbitrator I cant see anything on the usb device from esxi itself. Also it seems ESXi has no real USB support aside from passthrough. It doesnt even have 'mount'.
|
#
?
Apr 4, 2017 15:53
|
|
|
|
| # ? May 23, 2024 09:55 |
|
|
I don't know anything about VSA, but is an iSCSI/NFS NAS an option? Anytime I am dealing with something like you are describing I use a NAS instead of a USB drive.
|
|
#
?
Apr 4, 2017 15:59
|
|
|
We have a couple CentOS 7 VMs in Vsphere that seem to have poo poo themselves after cloning. The original parent VMs are fine but we can't get the clones to perform worth a damned - it looks like they can't get any CPU time (running "time iptables -L -v" takes 30 seconds of wall-clock time and gets 0.003 seconds of CPU time). However, the IT guys are telling me that they don't have any limits set in the VM consoles. They also tell me they're not seeing any signs of I/O latency on the SAN. They did reduce the number of cores the VM had when they cloned it but that shouldn't do anything (it's been restarted numerous times since changing the number of cores). We're all out of ideas on what might be causing this, any ideas?
|
|
#
?
Apr 4, 2017 16:10
|
|
|
Internet Explorer posted:I don't know anything about VSA, but is an iSCSI/NFS NAS an option? Anytime I am dealing with something like you are describing I use a NAS instead of a USB drive. SFTP from the cheapest Synology shitbox?
|
|
#
?
Apr 4, 2017 16:17
|
|
|
What's the COSTOP /WAIT/READY on the VMs and host(s)?
|
|
#
?
Apr 4, 2017 16:18
|
|
|
thebigcow posted:SFTP from the cheapest Synology shitbox? I don't know if SFTP is necessary if you can make a VMFS partition on an iSCSI NAS, but again, I haven't worked with VSA.
|
|
#
?
Apr 4, 2017 16:19
|
|
|
Internet Explorer posted:I don't know anything about VSA, but is an iSCSI/NFS NAS an option? Anytime I am dealing with something like you are describing I use a NAS instead of a USB drive.
|
|
#
?
Apr 4, 2017 16:23
|
|
|
The cheap WD NAS boxes support NFS, which surprised me. https://www.wdc.com/en-gb/products/network-attached-storage/my-cloud-expert-series-ex2-ultra.html
|
|
#
?
Apr 4, 2017 18:47
|
|
|
Can confirm, cheap WD boxes work in a pinch. I've had terrible experiences with Seagate ones though for what it's worth.
|
|
#
?
Apr 4, 2017 19:00
|
|
|
The advantage of SFTP or SCP or whatever would be not having to set up a datastore first. I don't know which is less scripting work in the end.
|
|
#
?
Apr 4, 2017 19:19
|
|
|
Someone created a 10TB VM on a datastore just big enough to fit it, took a Snapshot then started making copies of 4TB of files. Only on our testing cluster but still 
|
|
#
?
Apr 5, 2017 07:16
|
|
|
We'd just upgraded to vSphere 6.5 and we're in the process of moving NetBackup jobs to vDP, and then today VMware says they're getting rid of vDP after 6.5  https://www.emc.com/en-us/microsites/offer-programs/vdp-eoa.htm
|
|
#
?
Apr 5, 2017 22:30
|
|
|
anthonypants posted:We'd just upgraded to vSphere 6.5 and we're in the process of moving NetBackup jobs to vDP, and then today VMware says they're getting rid of vDP after 6.5 Not surprising, it never really took off.
|
|
#
?
Apr 5, 2017 23:27
|
|
|
anthonypants posted:We'd just upgraded to vSphere 6.5 and we're in the process of moving NetBackup jobs to vDP, and then today VMware says they're getting rid of vDP after 6.5 aaand now the customer who wanted vmware data protection last year gets to receive an email from me!
|
|
#
?
Apr 6, 2017 00:14
|
|
|
big money big clit posted:Not surprising, it never really took off.
|
|
#
?
Apr 6, 2017 00:20
|
|
|
anthonypants posted:We've had a DataDomain backup appliance forever, and vDP is free, but we've been using NetBackup because of inertia, and I'm worried that we'll go back to NetBackup instead of like, Veeam. If it's any consolation, VDP wasn't a particularly good product, and we've had customers that ran into significant issues with it, so maybe it's for the best.
|
|
#
?
Apr 6, 2017 21:28
|
|
|
big money big clit posted:If it's any consolation, VDP wasn't a particularly good product, and we've had customers that ran into significant issues with it, so maybe it's for the best.
|
|
#
?
Apr 6, 2017 21:30
|
|
|
It's a Symantec product, it's definitely worse.
|
|
#
?
Apr 7, 2017 00:57
|
|
|
VDP is just Avamar in different clothes. That's not much better, and I loving hate Netbackup.
|
|
#
?
Apr 7, 2017 01:35
|
|
|
devmd01 posted:Symantec product I was hoping the were gong out of business after the cert thing
|
|
#
?
Apr 7, 2017 16:32
|
|
|
Veeam is pretty awesome. We had Commvault and Veeam is just so nice compared to the complexity of Commvault. Spinning up machines in a lab for upgrading or testing purposes straight from the NetApp snapshots is pretty cool.
|
|
#
?
Apr 7, 2017 19:11
|
|
|
Rubrik is currently my favorite backup product out there. It's still somewhat limited though, and not cheap. VEEAM is great up to a certain size, but it has issues with scale and self-service/automation.
|
|
#
?
Apr 7, 2017 19:20
|
|
|
Scale as in backup sizes, VM count, host count, Veeam proxy count etc. or all of the above? How large are we talking here?
|
|
#
?
Apr 7, 2017 19:24
|
|
|
Thanks Ants posted:Scale as in backup sizes, VM count, host count, Veeam proxy count etc. or all of the above? How large are we talking here? Scale as in host/VM count. We're talking hundreds/thousands of hosts and thousands of VMs. For tens of hosts and a few hundreds of VMs it's fine, but things like proxy availability, repository performance, Veeam server perforance, etc get much harder to architect when you're trying to do dozens of even hundreds of concurrent backups.
|
|
#
?
Apr 7, 2017 19:30
|
|
|
We're seriously looking at Cohesity to replace EMC Notworker the next fiscal year, anyone have experience with that in a mostly virtualized environment? We will be migrating some workloads to the cloud within the next year, and we like everything about Cohesity so far.
|
|
#
?
Apr 7, 2017 19:58
|
|
|
big money big clit posted:Rubrik is currently my favorite backup product out there. It's still somewhat limited though, and not cheap. VEEAM is great up to a certain size, but it has issues with scale and self-service/automation. I agree 100% on both counts. We went from Veeam (which didn't scale for poo poo) to Rubrik and it's been fantastic. devmd01 posted:We're seriously looking at Cohesity to replace EMC Notworker the next fiscal year, anyone have experience with that in a mostly virtualized environment? Maybe worth looking at Rubrik too. Cohesity wasn't quite there yet when we ended up going Rubrik, but it certainly looked like a good solution.
|
|
#
?
Apr 7, 2017 21:40
|
|
|
Are the people hitting scalability issues with Veeam taking VMware snapshots and then shifting those to storage, or are you still going to have issues with database scale even if you're using storage snapshot integration?
|
|
#
?
Apr 7, 2017 21:51
|
|
|
My biggest beef with Veeam, as I can't speak to scale, is the way it handles archival/GFS backups. Doing something like 4 weekly / 12 monthly / 7 yearly archival backups takes waaaaay too much storage and they could handle it better. Otherwise in our small environment it has been great.
|
|
#
?
Apr 7, 2017 22:10
|
|
|
devmd01 posted:We're seriously looking at Cohesity to replace EMC Notworker the next fiscal year, anyone have experience with that in a mostly virtualized environment? We just partnered with Cohesity, so I don't have a ton of experience with them, but it's very similar to Rubrik and I've heard good things from a few colleagues. The architecture seems fine, it's just so new it's hard to say how it really works. Thanks Ants posted:Are the people hitting scalability issues with Veeam taking VMware snapshots and then shifting those to storage, or are you still going to have issues with database scale even if you're using storage snapshot integration? Depends on what you're doing with the storage offload. If you're just using VEEAM as a scheduler for storage snapshot and replication and recovery from storage snapshots then it's better. If you're still using VEEAM as a data mover you can still run into issues with having to deploy a ton of proxies to keep all of that data moving fast enough to meet backup windows. The big benefit of Rubrik and Cohesity from a scale perspective is that the ingest rate and concurrency naturally scale up as you add capacity. It also doesn't consume hypervisor compute resources like VEEAM proxies.
|
|
#
?
Apr 7, 2017 22:29
|
|
|
Is there really no way to do vCenter failover without a loadbalancer? If I have two sites, at which site would I put the loadbalancer?
anthonypants fucked around with this message at 00:04 on Apr 8, 2017 |
|
#
?
Apr 7, 2017 23:59
|
|
|
There's the HA option in 6.5 if you use the VCSA https://pubs.vmware.com/vsphere-65/index.jsp#com.vmware.vsphere.avail.doc/GUID-33AC12C8-EEB7-422D-831B-B1B5A7FECC44.html
|
|
#
?
Apr 8, 2017 00:23
|
|
|
anthonypants posted:Is there really no way to do vCenter failover without a loadbalancer? If I have two sites, at which site would I put the loadbalancer? The load balancer is just for the psc portion, not vcenter as well. If one PSC goes down you can just manually repoint vcenter to the replicated PSC to avoid the load balancer. Or use DNS to mimic a load balancer failover. But if the vcenter server goes down you've got to restore it from backup, or a replica, or otherwise get the VMware back up and running. Or use the windows vcenter with windows failover clustering for true failover. That's for 6.0. As mentioned, 6.5 has native HA for the vcsa.
|
|
#
?
Apr 8, 2017 01:41
|
|
|
I was loving around with our PSC and I think changing the timezone from UTC to PDT broke everything and now vpxd is complaining about certificate and the certificate-manager service won't start. How bad did I gently caress up
|
|
#
?
Apr 8, 2017 03:52
|
|
|
anthonypants posted:I was loving around with our PSC and I think changing the timezone from UTC to PDT broke everything and now vpxd is complaining about certificate and the certificate-manager service won't start. How bad did I gently caress up Give it 7 hours and it'll probably start working again.
|
|
#
?
Apr 8, 2017 04:48
|
|
|
theperminator posted:Give it 7 hours and it'll probably start working again.
|
|
#
?
Apr 8, 2017 17:43
|
|
|
anthonypants posted:I was loving around with our PSC and I think changing the timezone from UTC to PDT broke everything and now vpxd is complaining about certificate and the certificate-manager service won't start. How bad did I gently caress up
|
|
#
?
Apr 10, 2017 17:07
|
|
|
We have a XenServer host running some Oracle DR VMs, for inexplicable reasons, and I just found out that XenServer has a XenCenter so I turned it on and it's actually kind of nice. But it complained that our license had expired, so I went to the website it told me to go to and filled out a form and got a new license, and they sent me the following email with the license key attached:
|
|
#
?
Apr 12, 2017 21:19
|
|
|
As a long-time XenServer admin, XenServer is garbage. What you're seeing is a reactivation of the free license, just like VMware you are supposed to renew it once a year.
|
|
#
?
Apr 12, 2017 21:28
|
|
|
Internet Explorer posted:XenServer is garbage. This is the truth.
|
|
#
?
Apr 12, 2017 21:56
|
|
|
|
| # ? May 23, 2024 09:55 |
|
|
And so is VMware support these days. We're going on a week with a case open because we can't delete any goddamn datastores out of our production cluster. I just had the case escalated because the guy that was assigned to our case isn't worth a drat.
|
|
#
?
Apr 12, 2017 22:20
|
|