I don't have the answer but I am interested in having it!
I have trespassed manualy in Navisphere a LUN allocated to ESX servers 3.0.2. I saw that the trespass was successful and without interruption of the throughput. Then I have manualy trespassed back the LUN and it was successful too.
But I guess the behaviour of this manual trespass may be different of the behaviour of a real trespass ?
Nope A tresspass is a tresspass whether it is manually or automatically and should be nondusruptive. It depends on your host and it's failover software whether you notice anything.
OK so Paul, could you confirm you use the ESX native failover ? If yes, please provide the result of "esxcfg-mpath -l" command to see if your using the default policy MRU and the status of each path. Which version of ESX server do you use ?
It's not really a problem as I can trespass the LUNs back manually but as we have so many it takes a while.
I was told by an EMC consultant that there might be a command that can trespaas VMware LUNs back automatically but he didn't know what it was off hand. It would be useful to know waht it is though. I don't know what verson of VMware we are using as I don't look after those servers but I know it is ESX3.x. I was also told in VMware ESX4 LUNs can be trespassed back automatically.
MRU simply uses the last active path that was working ok. So after a tresspass the new path will be the MRU.
However if you have 2 or more ESX hosts and you're not using the MRU setting, but fixed instead, you can imagine what happens when 1 LUN is being accessed by more than 1 host and 1 is using another path than the other: tresspasses !
Just wanted to address one of the original questions that went unanswered. OP had asked what the command to bulk trespass (back) was. I'm sure by now, they found the answer as I doubt anyone would continue to do it in the GUI. On a side note, if within in the "Trespassed LUNs" window itself in Navisphere, you could trespass (back) which wouldn't make it too bad to do from the GUI. The cli command is:
naviseccli -h trespass mine
naviseccli -h trespass mine
As the command parameters suggests, it steps through each LUN and makes sure that "default owner = current owner". You would run it for both SP A and SP B.
As for the supported CLARiiON fail-over modes and ESX PSP policies, there is a chart found in the "White Paper: EMC CLARiiON Integration with VMware ESX—Applied Technology paper" available in PowerLink":
Home > Solutions > Horizontal Solutions > VMware Resource Center > Solution Areas > Shared Storage Platforms > Product Specific Kit > CLARiiON
However, when setting to MRU or Round Robin (where supported), keep the following article from EMC in mind in regards to LUN trespassing:
emc232355: "Using the VMware ESX native multipath plug-in (NMP) with a CLARiiON array"
as well as the following article from VMware:
KB Article: 1021814 "LUNs on Active/Passive arrays may trespass when connected to a running ESX 4.0 host"
Where supported (see chart above), we are suggesting: CLARiiON fail-over mode 4 (ALUA) and ESX Round Robin. One only needs to refer to any of the popular blogs such as Chad Sakac's or Clinton's who both repeat the same. As the name suggests, you could also set the NMP to "Fixed", but there is some initial configuration to define the preferred paths. Furthermore, this could possibly be simply masking the problem as it trespasses back to the preferred path, but doesn't necessarily address the initial trespass due to inconsistent decision tree, inconsistent PSA (NMP or MPP), etc. Of course, instead of using any of the ESX native solutions you can instead consider EMC's (third-part) MPP solution: PowerPath/VE. A highly recommended article to answer why you might be finding an excess of trespassed LUNs in the first place, and how to properly configure your environment to minimize it, please read Clinton's post:
"Why do I have trespassed LUNs or all LUNs on one SP?"
In addition to the white paper above, have a look at the EMC® Host Connectivity Guide for VMware ESX Server
From PowerLink:
Home > Support > Technical Documentation and Advisories > Host Connectivity/HBAs > Installation/Configuration
Topics:
1) Multipathing in VMware ESX Server
2) CLARiiON fail-over modes
Finally, to take advantage of VAAI (vStorage APIs for Array Integration) available with ESX 4.1 and FLARE 30, this requires ALUA (CLARiiON fail-over mode of 4) to support the hardware acceleration features which is something to possibly look forward to.
I would make one recommendation - use FIXED policy instead of Round Robin - FIXED will failback in the event of a path lose and RR will not. This is only when using the ALUA (failover mode 4 on the CLARiiON) mode.
ovivier
142 Posts
0
April 25th, 2008 01:00
I have trespassed manualy in Navisphere a LUN allocated to ESX servers 3.0.2. I saw that the trespass was successful and without interruption of the throughput.
Then I have manualy trespassed back the LUN and it was successful too.
But I guess the behaviour of this manual trespass may be different of the behaviour of a real trespass ?
ovivier
142 Posts
0
April 25th, 2008 03:00
Paul, did you meet this trespass back problem ?
RRR
4 Operator
•
5.7K Posts
0
April 25th, 2008 03:00
A tresspass is a tresspass whether it is manually or automatically and should be nondusruptive. It depends on your host and it's failover software whether you notice anything.
ovivier
142 Posts
0
April 25th, 2008 03:00
If yes, please provide the result of "esxcfg-mpath -l" command to see if your using the default policy MRU and the status of each path.
Which version of ESX server do you use ?
OrangePaul
1 Rookie
•
34 Posts
0
April 28th, 2008 03:00
I was told by an EMC consultant that there might be a command that can trespaas VMware LUNs back automatically but he didn't know what it was off hand. It would be useful to know waht it is though. I don't know what verson of VMware we are using as I don't look after those servers but I know it is ESX3.x.
I was also told in VMware ESX4 LUNs can be trespassed back automatically.
RRR
4 Operator
•
5.7K Posts
0
April 28th, 2008 05:00
OrangePaul
1 Rookie
•
34 Posts
0
May 22nd, 2008 04:00
esxcfg-mpath -l will give you a list of paths to LUNs and what¿s preferred and what¿s active
[root@dev01 root]# esxcfg-mpath -l
Disk vmhba0:0:0 /dev/sda (139899MB) has 1 paths and policy of Fixed
Local 4:0.0 vmhba0:0:0 On active preferred
Disk vmhba1:0:0 /dev/sdb (230400MB) has 2 paths and policy of Most Recently Used
FC 15:4.0 210000e08b9c1e11<->200400a0b829871d vmhba1:0:0 Standby active preferred
FC 15:4.0 210000e08b9c1e11<->200500a0b829871e vmhba1:1:0 On
Disk vmhba1:0:1 /dev/sdc (230400MB) has 2 paths and policy of Most Recently Used
FC 15:4.0 210000e08b9c1e11<->200400a0b829871d vmhba1:0:1 On active preferred
FC 15:4.0 210000e08b9c1e11<->200500a0b829871e vmhba1:1:1 Standby
Disk vmhba1:0:2 /dev/sdd (230400MB) has 2 paths and policy of Most Recently Used
FC 15:4.0 210000e08b9c1e11<->200400a0b829871d vmhba1:0:2 Standby active preferred
FC 15:4.0 210000e08b9c1e11<->200500a0b829871e vmhba1:1:2 On
But what if you want to set a different preferred path?
[root@esxdev01 root]# esxcfg-mpath ¿preferred ¿path=vmhba1:1:0 ¿lun=vmhba1:0:0
Setting vmhba1:0:0 ¿ vmhba1:1:0 as preferred path
[root@esxdev01 root]# esxcfg-mpath ¿path=vmhba1:1:0 ¿lun=vmhba1:0:0 ¿state=on
Setting vmhba1:0:0 ¿ vmhba1:1:0 state to on
Now you can see the results:
[root@esxdev01 root]# esxcfg-mpath -l
Disk vmhba0:0:0 /dev/sda (139899MB) has 1 paths and policy of Fixed
Local 4:0.0 vmhba0:0:0 On active preferred
Disk vmhba1:0:0 /dev/sdb (230400MB) has 2 paths and policy of Most Recently Used
FC 15:4.0 210000e08b9c1e11<->200400a0b829871d vmhba1:0:0 Standby active
FC 15:4.0 210000e08b9c1e11<->200500a0b829871e vmhba1:1:0 On preferred
Disk vmhba1:0:1 /dev/sdc (230400MB) has 2 paths and policy of Most Recently Used
FC 15:4.0 210000e08b9c1e11<->200400a0b829871d vmhba1:0:1 On active preferred
FC 15:4.0 210000e08b9c1e11<->200500a0b829871e vmhba1:1:1 Standby
Disk vmhba1:0:2 /dev/sdd (230400MB) has 2 paths and policy of Most Recently Used
FC 15:4.0 210000e08b9c1e11<->200400a0b829871d vmhba1:0:2 Standby active preferred
FC 15:4.0 210000e08b9c1e11<->200500a0b829871e vmhba1:1:2 On
To make the preferred path active now. Run esxcfg-rescan followed by the vmhba you want to scan.
OrangePaul
1 Rookie
•
34 Posts
0
June 4th, 2008 05:00
RRR
4 Operator
•
5.7K Posts
0
June 4th, 2008 05:00
However if you have 2 or more ESX hosts and you're not using the MRU setting, but fixed instead, you can imagine what happens when 1 LUN is being accessed by more than 1 host and 1 is using another path than the other: tresspasses !
Message was edited by:
RRR
On Clariion only MRU is supported
kelleg
4 Operator
•
4.5K Posts
0
June 4th, 2008 08:00
glen
StuartA461
1 Rookie
•
63 Posts
0
February 1st, 2011 06:00
When connecting to a CLARiiON with ESX 4.5, what VMware Failover Policies are supported, or recommended?
kelleg
4 Operator
•
4.5K Posts
0
February 1st, 2011 10:00
Please see my answer in thread:
https://community.emc.com/thread/116546?tstart=0
glen
christopher_ime
4 Operator
•
2K Posts
0
February 1st, 2011 19:00
Just wanted to address one of the original questions that went unanswered. OP had asked what the command to bulk trespass (back) was. I'm sure by now, they found the answer as I doubt anyone would continue to do it in the GUI. On a side note, if within in the "Trespassed LUNs" window itself in Navisphere, you could trespass (back) which wouldn't make it too bad to do from the GUI. The cli command is:
naviseccli -h trespass mine
naviseccli -h trespass mine
As the command parameters suggests, it steps through each LUN and makes sure that "default owner = current owner". You would run it for both SP A and SP B.
As for the supported CLARiiON fail-over modes and ESX PSP policies, there is a chart found in the "White Paper: EMC CLARiiON Integration with VMware ESX—Applied Technology paper" available in PowerLink":
Home > Solutions > Horizontal Solutions > VMware Resource Center > Solution Areas > Shared Storage Platforms > Product Specific Kit > CLARiiON
However, when setting to MRU or Round Robin (where supported), keep the following article from EMC in mind in regards to LUN trespassing:
emc232355: "Using the VMware ESX native multipath plug-in (NMP) with a CLARiiON array"
as well as the following article from VMware:
KB Article: 1021814
"LUNs on Active/Passive arrays may trespass when connected to a running ESX 4.0 host"
Where supported (see chart above), we are suggesting: CLARiiON fail-over mode 4 (ALUA) and ESX Round Robin. One only needs to refer to any of the popular blogs such as Chad Sakac's or Clinton's who both repeat the same. As the name suggests, you could also set the NMP to "Fixed", but there is some initial configuration to define the preferred paths. Furthermore, this could possibly be simply masking the problem as it trespasses back to the preferred path, but doesn't necessarily address the initial trespass due to inconsistent decision tree, inconsistent PSA (NMP or MPP), etc. Of course, instead of using any of the ESX native solutions you can instead consider EMC's (third-part) MPP solution: PowerPath/VE. A highly recommended article to answer why you might be finding an excess of trespassed LUNs in the first place, and how to properly configure your environment to minimize it, please read Clinton's post:
"Why do I have trespassed LUNs or all LUNs on one SP?"
https://community.emc.com/thread/114381?tstart=30
In addition to the white paper above, have a look at the EMC® Host Connectivity Guide for VMware ESX Server
From PowerLink:
Home > Support > Technical Documentation and Advisories > Host Connectivity/HBAs > Installation/Configuration
Topics:
1) Multipathing in VMware ESX Server
2) CLARiiON fail-over modes
Finally, to take advantage of VAAI (vStorage APIs for Array Integration) available with ESX 4.1 and FLARE 30, this requires ALUA (CLARiiON fail-over mode of 4) to support the hardware acceleration features which is something to possibly look forward to.
http://virtualgeek.typepad.com/virtual_geek/2010/07/vsphere-41---what-do-the-vstorage-apis-for-array-integration-mean-to-you.html
kelleg
4 Operator
•
4.5K Posts
0
February 2nd, 2011 10:00
I would make one recommendation - use FIXED policy instead of Round Robin - FIXED will failback in the event of a path lose and RR will not. This is only when using the ALUA (failover mode 4 on the CLARiiON) mode.
glen
RRR
4 Operator
•
5.7K Posts
0
February 3rd, 2011 01:00
Ok, so use fixed, mode 4 and ALUA ?