Register a SA Forums Account here!
JOINING THE SA FORUMS WILL REMOVE THIS BIG AD, THE ANNOYING UNDERLINED ADS, AND STUPID INTERSTITIAL ADS!!!

You can: log in, read the tech support FAQ, or request your lost password. This dumb message (and those ads) will appear on every screen until you register! Get rid of this crap by registering your own SA Forums Account and joining roughly 150,000 Goons, for the one-time price of $9.95! We charge money because it costs us $3,400 per month for bandwidth bills alone, and since we don't believe in shoving popup ads to our registered users, we try to make the money back through forum registrations.
«282 »
  • Post
  • Reply
MC Fruit Stripe
Nov 26, 2002

around and around we go


Tev posted:

You can just drop the new license in when you renew your VMUG Advantage and be good for another year.
Okay cool, and thank you. I was taking the word 'download' perhaps a little too literally, as though you had to reinstall everything perhaps as a security measure against using these licenses in production.

MC Fruit Stripe fucked around with this message at Oct 13, 2016 around 12:51

Adbot
ADBOT LOVES YOU

Spring Heeled Jack
Feb 25, 2007


VMware leaked their AWS partnership announcement a little early: https://webcache.googleusercontent....n&ct=clnk&gl=us

I was googling to find when the press conf was scheduled and a link to their blog showed up! Thankfully google cache grabbed a copy.

Thanks Ants
May 21, 2004

Bless You Ants, Blants



Fun Shoe

Did vCloud Air ever get off the ground?

YOLOsubmarine
Oct 19, 2004

When asked which Pokemon he evolved into, Kamara pauses.

"Motherfucking, what's that big dragon shit? That orange motherfucker. Charizard."


Thanks Ants posted:

Did vCloud Air ever get off the ground?

Not especially, hence their partnerships around vCloud Foundation with IBM and now AWS.

Thanks Ants
May 21, 2004

Bless You Ants, Blants



Fun Shoe

Yeah that's what I heard as well, the entire UK deployed infrastructure was a few racks as of a year ago.

Maneki Neko
Oct 27, 2000



Thanks Ants posted:

Did vCloud Air ever get off the ground?

They pretty much said "yeah that poo poo ain't working, we'll help with your public cloud stuff" at this years vmworld.

Thanks Ants
May 21, 2004

Bless You Ants, Blants



Fun Shoe

It's announced now

https://www.vmware.com/cloud-servic...-cloud-aws.html

Tab8715
May 20, 2006

NOBODY WANTS TO HEAR YOUR NIGHTCORE GARBAGE, MORON


Is this similar to Azure Site Recovery or this and a whole lot more?

keykey
Mar 28, 2003

     


Looks like Hyper-v server 2016 is up on MS download page: https://www.microsoft.com/en-us/eva...ows-server-2016

it says evaluation but doesn't give a time limit, much like Hyper-v 2012 r2, except it states unlimited. Is this the final build or is it still a technical preview? Also, I would assume it's an unlimited "evaluation"?

adorai
Nov 2, 2002

10/27/04 Never forget

Grimey Drawer

edit: wrong thread.

HPL
Aug 28, 2002

Worst case scenario.

So kind of shooting in the dark here, but there's a 2012R2 terminal server farm at work that uses profile disks and it has been flaky as heck, with peoples VHDs occasionally not detaching when they log off and crazy indexing problems with Outlook causing message search issues. Anyone ever get profile disks working nicely?

theperminator
Sep 16, 2009

by Smythe


Fun Shoe

Has anyone had any issues with vMotioning a Database?

In all the years I've been using vSphere we've never had an issue with vMotion causing any data issues, at my current workplace our DRS threshold is set quite conservatively and we've recently had a MySQL Consultant tell us we should not vMotion our Databases at all. But obviously we do vMotion the database hosts whenever we patch our esxi hosts or perform maintenance and have never had any issues before.
I said that was bullshit, at my last place we had plenty of them with DRS in Full-Auto aggressively migrating databases & other workloads around without any issues.

Recently we did a test, we simply vmotioned our database slave to every host in our cluster, while this was happening we got relay-log corruption and slaving has stopped.
I'm scratching my head trying to figure out what happened, at the same time we were doing this it moved another DB Slave around to maintain affinity rules, the other VM didn't have any issues at all.

I Find it hard to believe that vMotion would be a risky thing, it's one of the selling points of the software and it's something we've done plenty of before but is only suddenly now causing issues?

adorai
Nov 2, 2002

10/27/04 Never forget

Grimey Drawer

Vmotion has to pause io for short periods. It should not affect normal db activity, but clustered or slaved activity could theoretically be impacted by the increased latency.

Vulture Culture
Jul 14, 2003

I was never enjoying it. I only eat it for the nutrients.


I'd be really surprised if I/O paused long enough to present a problem on its own, but especially in robust clustering systems, you can drop heartbeats for long enough to trigger failovers while the checkpoints converge on the destination host and the network cuts over. A rapid failover/fail-back is the most likely culprit when looking for the source of log corruption, but you wouldn't see that when moving the slave, only the master. I'm confused myself.

theperminator
Sep 16, 2009

by Smythe


Fun Shoe

Still scratching my head too, there was no failover actions taken, the migration completed and the VM did not pause during the migration.

Looking at the binlog, there is 2.5MB of zero bytes starting at the query it has failed on, the binlogs continue after that as normal (the slave was lagging by a few hours when this happened, the IO thread is ahead of the SQL Thread)
FSCK reports no errors on the filesystem either.

It's like a file buffer was zero'ed somehow during the process of migration, I would expect the vmotion process to do some kind of checksum but maybe not?

keykey
Mar 28, 2003

     


Out of curiosity, is there actually any full featured free hyper-v management software that could install on a hyper-v server so instead of having to remote administer from a win7/10 box you could just do it all on a free edition hyper-v 2012/2016 box? I've tried 5nine manager, but to export/etc.. there is an incurred cost.

Potato Salad
Oct 23, 2014

Nobody Cares



Tortured By Flan

...powershell?

I'm kidding, I'm kidding.

Jeoh
Jul 20, 2010



keykey posted:

Out of curiosity, is there actually any full featured free hyper-v management software that could install on a hyper-v server so instead of having to remote administer from a win7/10 box you could just do it all on a free edition hyper-v 2012/2016 box? I've tried 5nine manager, but to export/etc.. there is an incurred cost.

No, that's why it's free.

e: Nested hypervisors all the way down!

jre
Sep 2, 2011

To the cloud ?





theperminator posted:

Still scratching my head too, there was no failover actions taken, the migration completed and the VM did not pause during the migration.

Looking at the binlog, there is 2.5MB of zero bytes starting at the query it has failed on, the binlogs continue after that as normal (the slave was lagging by a few hours when this happened, the IO thread is ahead of the SQL Thread)
FSCK reports no errors on the filesystem either.

It's like a file buffer was zero'ed somehow during the process of migration, I would expect the vmotion process to do some kind of checksum but maybe not?

Do you have sync_binlog =0 set on the affected dbs my.conf ?

theperminator
Sep 16, 2009

by Smythe


Fun Shoe

jre posted:

Do you have sync_binlog =0 set on the affected dbs my.conf ?

Just checked, we are running an old version where this is the default, confirmed with a "show global variables"
It might be worthwhile for us to modify this, I notice that on 5.7.7 and up the new default is 1

Still weird though, the vmotion shouldn't impact the guest to my knowledge but I guess there's a difference between theory and practice.

keykey
Mar 28, 2003

     


Jeoh posted:

No, that's why it's free.

e: Nested hypervisors all the way down!

After having to use another machine for the past week for hyper-v manager, I did some digging and found my answer today. 5nine manager free version does everything I want it to except export/import functions, because you gotta pay for that and I'm not about paying over $100 for something I'm screwing around with at home. Another utility pshvm30 fills in the blanks of being able to import/export/do everything else 5nine manager does albeit in a seriously clunky way besides the import/export functions. Between those 2, you can get a tarted up enough server to do all the functions you need while still being absolutely free.

MC Fruit Stripe
Nov 26, 2002

around and around we go


Silly problem with VMware Workstation on Windows 10 which I never encountered on 7 (I skipped 8).

With UAC *on* but at its lowest setting, I can't modify virtual networks. When I open the virtual network editor, I see the 3 default VM networks, but can't change anything. I have to click the elevate 'change settings' button, at which point it re-opens and I can see the *4* VM networks I've configured. This also means that if I want to use my VM networks, I have to run VMware workstation as administrator. If I do THAT, then VMware workstation doesn't have access to mapped drives because they're all running under my user account. If I disable UAC completely, first of all that's not what I'm going to be doing, but just playing along, it does fix the issue but then I can't run sandbox apps such as Calc or Edge anymore.

Has anyone seen this behavior? I googled "virtual network editor" "administrator privileges" and a few other combos and apparently I am the first person ever to encounter this. I looked for the config files, and I found the 4 VM administrator one hiding under C:\ProgramData\VMware, but a full search of my C drive turned up no other copies.



e: See what happened here? I made a post and instantly found the solution. That's how these things work, after an hour of digging, just ask and you'll find it yourself. EnableLinkedConnections registry key set me up, allowing my elevated user access to the mapped drives.

MC Fruit Stripe fucked around with this message at Oct 19, 2016 around 19:56

Martytoof
Feb 25, 2003

It's called a hassle, sweetheart..



In TYOOL 2016 why doesn't vSphere have a way of reconnecting to an NFS datastore without me having to esxcli storage nfs delete and re-add it? :|

I mean this is just my homelab so it's not critical but c'moooon

Thanks Ants
May 21, 2004

Bless You Ants, Blants



Fun Shoe

I can reboot an NFS datastore and vSphere (5.5) will reconnect when it comes back up again. Didn't do any configuration to make this happen as far as I can remember.

Martytoof
Feb 25, 2003

It's called a hassle, sweetheart..



Thanks Ants posted:

I can reboot an NFS datastore and vSphere (5.5) will reconnect when it comes back up again. Didn't do any configuration to make this happen as far as I can remember.

Hm. Weird. I rebooted my NAS and the NFS datastore won't come back up unless I esxcli storage nfs delete and re-add :[

I guess I could be an edge case.

anthonypants
May 6, 2007

by Nyc_Tattoo


Dinosaur Gum

You could use PowerCLI

evil_bunnY
Apr 2, 2003



Martytoof posted:

In TYOOL 2016 why doesn't vSphere have a way of reconnecting to an NFS datastore without me having to esxcli storage nfs delete and re-add it? :|

I mean this is just my homelab so it's not critical but c'moooon
Your poo poo's broke homey

theperminator
Sep 16, 2009

by Smythe


Fun Shoe

Martytoof posted:

In TYOOL 2016 why doesn't vSphere have a way of reconnecting to an NFS datastore without me having to esxcli storage nfs delete and re-add it? :|

I mean this is just my homelab so it's not critical but c'moooon

Which version of ESXi are you running? I used to have this problem a while back with 5.5 but I think one of the updates has fixed it

edit:
https://www.vmguru.com/2015/12/best...vmware-with-nfs

quote:

Make sure your VMware environment is updated and runs on patch 5 or newer if you have ESXi 5.5 and on Update 1a if you run with ESXi 6.0
Which version of VMware are you running?
Running ESXi 5.5 below Patch 5? Than you are running on a build number lower than 2718055 and it is time to upgrade now!
We have seen several virtual environments that ran below patch 5 and had strange and unpredictable behavior from time to time. Like disconnects (offline, greyed out) which could lead to an all paths down (APD) event (for 45 minutes) or even permanent device loss (PDL) or management interfaces which freeze for several minutes or getting stuck majorly. After updating those infrastructures to newer builds you could restart an ESXi server and datastores where almost instantly back online and visible. Also managing storage in vCenter server becomes more fluid than before. And a rescan completes about 3 to 4 times faster.

theperminator fucked around with this message at Oct 23, 2016 around 06:11

Martytoof
Feb 25, 2003

It's called a hassle, sweetheart..



theperminator posted:

Which version of ESXi are you running? I used to have this problem a while back with 5.5 but I think one of the updates has fixed it

edit:
https://www.vmguru.com/2015/12/best...vmware-with-nfs

It's a homelab so I'm probably lax on updates because it's "working" right now, but I'm definitely past 1a as I've done an update to "latest" this year.

I think part of the problem is that vSphere thought the share was "in use" when I tried to delete it in the UI but when I deleted it via esxcli it didn't peep.

I'm perfectly happy assuming it's something wrong with my setup for what it's worth, I've just had this happen every time I reboot my NFS share (which I do like once every few months for reasons).

Kachunkachunk
Jun 6, 2011


MC Fruit Stripe posted:

Silly problem with VMware Workstation on Windows 10 which I never encountered on 7 (I skipped 8).

With UAC *on* but at its lowest setting, I can't modify virtual networks. When I open the virtual network editor, I see the 3 default VM networks, but can't change anything. I have to click the elevate 'change settings' button, at which point it re-opens and I can see the *4* VM networks I've configured. This also means that if I want to use my VM networks, I have to run VMware workstation as administrator. If I do THAT, then VMware workstation doesn't have access to mapped drives because they're all running under my user account. If I disable UAC completely, first of all that's not what I'm going to be doing, but just playing along, it does fix the issue but then I can't run sandbox apps such as Calc or Edge anymore.

Has anyone seen this behavior? I googled "virtual network editor" "administrator privileges" and a few other combos and apparently I am the first person ever to encounter this. I looked for the config files, and I found the 4 VM administrator one hiding under C:\ProgramData\VMware, but a full search of my C drive turned up no other copies.



e: See what happened here? I made a post and instantly found the solution. That's how these things work, after an hour of digging, just ask and you'll find it yourself. EnableLinkedConnections registry key set me up, allowing my elevated user access to the mapped drives.
Well that's weird, I guess if nobody can suggest anything, I'm earmarking your post to check into it more tomorrow and/or get a bug filed. For what it's worth anyway!
I don't really have a suggestion, though, sorry. I'm one of those idjits that disables UAC because... bareback, baby!

MC Fruit Stripe
Nov 26, 2002

around and around we go


Kachunkachunk posted:

Well that's weird, I guess if nobody can suggest anything, I'm earmarking your post to check into it more tomorrow and/or get a bug filed. For what it's worth anyway!
I don't really have a suggestion, though, sorry. I'm one of those idjits that disables UAC because... bareback, baby!
I never did figure it out and just need to remember to start Workstation with elevated permissions. I paired that with the registry key EnableLinkedConnections, which allows Workstation which is now running as Administrator rather than my account, to access network shares. Definitely a weird one.

Sepist
Dec 25, 2005

FUCK BITCHES, ROUTE PACKETS


Gravy Boat 2k

Not sure the best way to do this. I have 2 hosted Ubunutu VM's that need to be migrated to my internal ESXi server. I don't have the ability to get an ova of the existing servers and I am using a bunch of modified kernels that I don't remember how I got to where they are today, so I'm trying to avoid starting from fresh. Is there any kind of linux based app will let me basically mirror the entire host and unpack it in it's new VM?

YOLOsubmarine
Oct 19, 2004

When asked which Pokemon he evolved into, Kamara pauses.

"Motherfucking, what's that big dragon shit? That orange motherfucker. Charizard."


Sepist posted:

Not sure the best way to do this. I have 2 hosted Ubunutu VM's that need to be migrated to my internal ESXi server. I don't have the ability to get an ova of the existing servers and I am using a bunch of modified kernels that I don't remember how I got to where they are today, so I'm trying to avoid starting from fresh. Is there any kind of linux based app will let me basically mirror the entire host and unpack it in it's new VM?

VMware converter?

milk milk lemonade
Jul 29, 2016


Uhhh what the christ kind of bug causes VSphere 6.0 to report "this VM cannot be found" when you go to power a VM back on.

Not that I'm worried about a clients production server or anything....




at least it's at a datacenter

edit: Obviously not really a big deal, but still jesus loving christ vmware give me a heart attack why dont you

milk milk lemonade fucked around with this message at Nov 4, 2016 around 02:06

HPL
Aug 28, 2002

Worst case scenario.

Finally put together my Xeon E5-2670 workstation with 64GB of RAM. Figured I'd give oVirt 4 a go with CentOS 7. Wow, oVirt installation sure has improved since the 3.x days. I was actually able to get it up and running.

evol262
Nov 30, 2010
#!/usr/bin/perl

4.1 lands soon, which will use ELK for metrics. oVirt Node was also rewritten from the ground up for 4.0, so most of the legacy problems are gone...

Martytoof
Feb 25, 2003

It's called a hassle, sweetheart..



Really wish I had a reason to switch to ovirt. Have no need for vsphere but right now everything is just running for the most part.

I'm building a malware lab so maybe I'll use oVirt for that though.

jaegerx
Sep 10, 2012

Maybe this post will get me on your ignore list!

Grimey Drawer

Sepist posted:

Not sure the best way to do this. I have 2 hosted Ubunutu VM's that need to be migrated to my internal ESXi server. I don't have the ability to get an ova of the existing servers and I am using a bunch of modified kernels that I don't remember how I got to where they are today, so I'm trying to avoid starting from fresh. Is there any kind of linux based app will let me basically mirror the entire host and unpack it in it's new VM?

Why are you modifying the kernel?

Sepist
Dec 25, 2005

FUCK BITCHES, ROUTE PACKETS


Gravy Boat 2k

It was so long ago that I don't remember the specific reason as to why. I can't remember if it was some hacky poo poo in order to run 30 router binaries at once or an issue I was having with a shared object.

Adbot
ADBOT LOVES YOU

BangersInMyKnickers
Nov 3, 2004

I have a thing for courageous dongles



I've been working with VMware support for a few months now on a really annoying issue with their LACP implementation. The first part is that vdSwitches with LACP default to long timeouts (90s) which is not exactly desirable for failed link detection on the storage fabric. The switches default to short timeouts (3s) and I'm forced to just live with the mismatch. You can manually configure each host to run the vdSwitch with short timeouts over SSH but there doesn't seem to be a way to change the vdSwitch definition so it resets back to long on reboot. This ends up manifesting itself at points in the night when disk traffic is extremely light and the host goes a few seconds without sending disk requests. The only thing moving at that point are the lacp pdus but the host is sending them out so slow that the switch thinks the link dies so it bounce, the host freaks out and re-establishes it, and your logs get flooded with nasty looking stuff.

That issue seems to be exacerbated by problem 2: Their lacp load-balancing implementation doesn't work. All outbound traffic from a single IP homes to a single link on the aggregate despite going to different target IPs and ports which their balancing algorithm should be dividing up. I get balanced traffic back in from the storage appliance IPs, but one of the two outbound links is completely dead, 0 packets/s and no amount of tweaking will get it to pass traffic. It fails over just fine, but the balancing is non-existent.

The CSR is hinting that 6.5 will fix the long timeout issue on the vdSwitches, but no word on the balancing algorithm. How the hell does this poo poo get past QA?

  • 1
  • 2
  • 3
  • 4
  • 5
  • Post
  • Reply
«282 »