LUNS are not visible with the latest created date

[S.S.]

Hi Experts,

Few days back i have create a 100gb vdisk at EVA and then created vg, filesystems.

After creation of VG at EVA i have presented the LUNS to the host then at the host i have executed the command #insf -e.
Later i checked the #autopath status there my new WWN and LUNS are visible.
After that i continued to create vg and filesystem.

I have a query that the new luns are visible under the new WWN in the #autopath status output whereas when i go to the
#cd /dev/dsk
#ll -lrt
These new LUNS are with the older date creation.

Kindly advice me anything i went wrong and why the LUNS are not showing with the latest date.

Thanks!

[Johnson Punniyalingam · Accepted solution]

#cd /dev/dsk
#ll -lrt
These new LUNS are with the older date creation

in this case can you try..

#ioscan -fn
#insf -e
#ioscan -fnC disk

#cd /dev/rdsk
#ls -lrt

[Kapil Jha]

I suppose there might have been some disk before with same name.
and when u created a new disk with same name it would just be there with old updated date.
No newly created device files untill previously you removed with rmsf or rebooted the mahine.

You can track it with the date you have that file.
BR,
Kapil+

[S.S.]

Hi Johnson,

Again i have followed your steps but the in the output of
#cd /dev/dsk
#ls -lrt
the LUNS are not updated with the latest date.

#cd /dev/rdsk
#ls -lrt
Only one LUN is showing the created date.
crw-r----- 1 bin sys 188 0x280700 Jan 27 09:39 c40t0d7

Kindly help me.

Thanks!

[Kapil Jha]

If everything is running fine......and you know which device is it exactly......
why you wane update files.....
if you want remove the device files with
#rmsf

then insf -eC disk
then new device files would be created.

Make sure you don delete wrong device file.

BR,
Kapil+

[R.K. #]

Hello S.S.

>>the LUNS are not updated with the latest date.
That is not required as already said.

Still, you can try rmsf:
# rmsf -H
# insf -eC disk

[Johnson Punniyalingam]

>>Only one LUN is showing the created date.
crw-r----- 1 bin sys 188 0x280700 Jan 27 09:39 c40t0d7<<<


so as per above post from you
can do check with "diskinfo" to check you LUN size is about "100GB" ?

#diskinfo -v /dev/rdsk/c40t0d7

[S.S.]

Hi Kapil,

May i know the device file dates will be changed after the reboot of the server?

If it can be done by rebooting the server then i will not do removing the device files because due to one of the activity we are shutting down the server in a short period.

otherwise i will try to follow your steps.

Kindly suggest me.

Thanks!

[R.K. #]

Hi Again,

>>the device file dates will be changed after the reboot of the server?
YEs

[S.S.]

Please find the output:

# diskinfo -v /dev/rdsk/c40t0d7
SCSI describe of /dev/rdsk/c40t0d7:
vendor: HP
product id: HSV210
type: direct access
size: 104857600 Kbytes
bytes per sector: 512
rev level: 6110
blocks per disk: 209715200
ISO version: 0
ECMA version: 0
ANSI version: 2
removable media: no
response format: 2
(Additional inquiry bytes: (32)35 (33)41 (34)37 (35)41 (36)55 (37)50 (38)41 (39)35 (40)34 (41)42 (42)35 (43)41 (44)37 (45)41 (46)55 (47)50 (48)39 (49)4f (50)59 (51)0 (52)0 (53)0 (54)62 (55)d (56)80 (57)13 (58)20 (59)8 (60)c0 (61)3 (62)0 (63)3 (64)24 (65)1 (66)60 (67)1 (68)c0 (69)0 (70)0 (71)0 (72)0 (73)0 (74)0 (75)0 (76)0 (77)0 (78)0 (79)0 (80)0 (81)0 (82)0 (83)0 (84)0 (85)0 (86)0 (87)0 (88)0 (89)0 (90)0 (91)91 (92)12 (93)0 (94)0 (95)0 (96)0 (97)0 (98)0 (99)0 (100)0 (101)0 (102)0 (103)0 (104)0 (105)0 (106)0 (107)0 (108)0 (109)0 (110)0 (111)0 (112)0 (113)0 (114)0 (115)0 (116)0 (117)0 (118)0 (119)0 (120)0 (121)0 (122)0 (123)c (124)80 (125)0 (126)0 (127)0 (128)0 (129)2 (130)0 (131)77 (132)fd (133)d0 (134)40 (135)77 (136)ff (137)8 (138)1c (139)0 (140)0 (141)0 (142)0 (143)0 (144)0 (145)0 (146)0 (147)0 (148)0 (149)0 (150)0 (151)48 (152)53 (153)56 (154)32 (155)31 (156)30 (157)20 (158)20 (159)20 (160)20 (161)20 (162)20 (163)20 (164)20 (165)20 (166)20 (167)0 (168)0 (169)0 (170)14 (171)c (172)7f (173)ff (174)ff (175)0 (176)0 (177)2 (178)0 (179)0 (180)0 (181)0 (182)0 (183)0 (184)0 (185)0 (186)0 (187)0 (188)0 (189)0 (190)0 (191)0 (192)0 (193)0 (194)0 (195)0 (196)0 (197)0 (198)0 (199)0 (200)0 (201)0 (202)0 (203)77 (204)fc (205)c0 (206)28 (207)0 (208)0 (209)0 (210)4 (211)0 (212)0 (213)0 (214)0 (215)0 (216)0 (217)0 (218)0 (219)0 (220)0 (221)0 (222)dd (223)40 (224)0 (225)18 (226)90 (227)77 (228)fd (229)67 (230)94 (231)0 (232)0 (233)0 (234)0 (235)0 (236)0 (237)3a (238)9f (239)c0 (240)20 (241)1a (242)ab (243)0 (244)0 (245)0 (246)0 )

[Kapil Jha]

Reboot would not change the date, it would just remove the orphan device files, and would create all the newly added hardware device file.

So if you already have, old files and its running the reboot wont help.

You can remove old device files[carefully], and run insf -eC disk
this would create new files.

BR,
kapil+

[S.S.]

Hi Kapil,

The following are the device files showing the older date:

brw-r----- 1 bin sys 31 0x0a0700 Oct 28 2007 c10t0d7
brw-r----- 1 bin sys 31 0x0c0700 Oct 28 2007 c12t0d7
brw-r----- 1 bin sys 31 0x0e0700 Oct 28 2007 c14t0d7
brw-r----- 1 bin sys 31 0x100700 Oct 28 2007 c16t0d7
brw-r----- 1 bin sys 31 0x120700 Oct 28 2007 c18t0d7
brw-r----- 1 bin sys 31 0x140700 Oct 28 2007 c20t0d7
brw-r----- 1 bin sys 31 0x160700 Oct 28 2007 c22t0d7
brw-r----- 1 bin sys 31 0x180700 Oct 28 2007 c24t0d7
brw-r----- 1 bin sys 31 0x1a0700 Oct 28 2007 c26t0d7
brw-r----- 1 bin sys 31 0x1c0700 Oct 28 2007 c28t0d7
brw-r----- 1 bin sys 31 0x1e0700 Oct 28 2007 c30t0d7
brw-r----- 1 bin sys 31 0x200700 Oct 28 2007 c32t0d7
brw-r----- 1 bin sys 31 0x220700 Oct 28 2007 c34t0d7
brw-r----- 1 bin sys 31 0x240700 Oct 28 2007 c36t0d7
brw-r----- 1 bin sys 31 0x260700 Oct 28 2007 c38t0d7
brw-r----- 1 bin sys 31 0x280700 Oct 28 2007 c40t0d7

and the device file /dev/dsk/c40t0d7 is used for the volume creation where as the other device files are the alternate links.

So, can i remove these device files?

[R.K. #]

You can very well remove all of these spl files and then run "insf -eC disk" to recreate all of them in one go.

Take ioscan output before starting.

[Kapil Jha]

Ok so you have so many alternate link...how many fc u have......??
so u can delete the below device files and run #insf -eC disk
it would create new device files and you would be able to use them again.

BR,
Kapil+

[R.K. #]

Just rebooted a 11.11 server & checked for time stamps:

BEFORE REBOOT:
-------------
/dev/dsk >ls -ltr
total 0
brw-r----- 1 bin sys 31 0x01f000 Feb 3 12:10 c1t15d0
brw-r----- 1 bin sys 31 0x03f000 Feb 3 12:10 c3t15d0


AFTER REBOOT:
-------------
/dev/dsk >ls -ltr
total 0
brw-r----- 1 bin sys 31 0x01f000 Feb 3 12:10 c1t15d0
brw-r----- 1 bin sys 31 0x03f000 Feb 3 12:10 c3t15d0

SO....
NO CHANGE IN TIME STAMPS

Thanks to Kapil.

[S.S.]

we have a 2node cluster.

On our production server all the filesystems are configured under the package config file. So for these vdisks we have presented 7 fcs (production+secondary server)

But i have created the filesystem which is not configured in the cluster so that i just presented the 4 fcds related to this server.

Totally for this one is 4 fcds.

For removing of these special device files which option should i prefer.

#rmsf -H
or
#rmsf

kindly suggest me.

[R.K. #]

Both will work:

#rmsf /dev/rdsk/cxtydz
or
#rmsf -H

[S.S.]

Kapil,

For the clustered filesystems we have 8fcds.

The one which i created is with 4 fcds. So in that case i should get 8 device files right?

[johnsonpk]

The one which i created is with 4 fcds. So in that case i should get 8 device files right?

Yes if you have eva with two host port per controller and all of them are zoned to your 4 hbas at server side ..

[Kapil Jha]

Yes I think u can delete all of them....and run insf -e
there would not be any issue.
Just if you wane cross check do a pvdisplay on each path to make sure its not part of any VG so that u wont lose any data...
(hope u don have VxVM cause pvdisplay do not work on vxvm)

BR,
Kapil+

[Johnson Punniyalingam]

Diskinfo -> output which you have posted confirm its "100GB" -> Lun

we can confirm "no issues" from From Storage side or from OS side,

Once ->Disk has been presented to Hosts-> ioscan ->scanning (device file are created) on that actual day, so the time stamp of the "devfiles" should be same.

[S.S.]

The device files are:
$ ll c*t0d7
brw-r----- 1 bin sys 31 0x0a0700 Oct 28 2007 c10t0d7
brw-r----- 1 bin sys 31 0x0c0700 Oct 28 2007 c12t0d7
brw-r----- 1 bin sys 31 0x0e0700 Oct 28 2007 c14t0d7
brw-r----- 1 bin sys 31 0x100700 Oct 28 2007 c16t0d7
brw-r----- 1 bin sys 31 0x120700 Oct 28 2007 c18t0d7
brw-r----- 1 bin sys 31 0x140700 Oct 28 2007 c20t0d7
brw-r----- 1 bin sys 31 0x160700 Oct 28 2007 c22t0d7
brw-r----- 1 bin sys 31 0x180700 Oct 28 2007 c24t0d7
brw-r----- 1 bin sys 31 0x1a0700 Oct 28 2007 c26t0d7
brw-r----- 1 bin sys 31 0x1c0700 Oct 28 2007 c28t0d7
brw-r----- 1 bin sys 31 0x1e0700 Oct 28 2007 c30t0d7
brw-r----- 1 bin sys 31 0x200700 Oct 28 2007 c32t0d7
brw-r----- 1 bin sys 31 0x220700 Oct 28 2007 c34t0d7
brw-r----- 1 bin sys 31 0x240700 Oct 28 2007 c36t0d7
brw-r----- 1 bin sys 31 0x260700 Oct 28 2007 c38t0d7
brw-r----- 1 bin sys 31 0x280700 Oct 28 2007 c40t0d7
$

I have removed the device file /dev/rdsk/c12t0d7
and did the #insf -eC disk but still showing the older date only.

Kapil,
You adviced me to remove all device files, i have created a vg with the device file /dev/rdsk/c40t0d7. then in that case can i delete this device file?

Can anyone clarify me that,
Actually i have created this new vg in the primary node of the cluster. All the filesystem in this node are configured in the cluster package but i didn't create this filesystem in the cluster package. I just created vg and a filesystem and added in the /etc/fstab file. If i do this will it make any problem during the reboot process?

Also kindly suggest me that am in the right track or i did anything wrong during the creation time of vdiks,vg and filesystem.
Will this device files problem make any issue during the reboot time?

Thank you all for helping me to troubleshoot this problem.

[Kapil Jha]

did you removed the disk device after creation of vg(leaving one device file), I am not sure but this can be a reason for older device name.

If you can delete vg and both /dev/dsk, /dev/rdsk device files and insf
Just to give a try....
else i suppose if everything is fine there is no need to trouble shoot.

You VGs would not create any problem as they would be mounted as local file system.
only problem can come via reboot is, if u have not put the good entries in fstab.

BR,
Kapil+

[S.S.]

Hi Kapil,

I have removed the all alternate links except leaving the vg creation device files and again executed the command #insf -eC disk then the time has been changed to today's date.

Thank you for resolving me this issue.

I would like to know one thing that, after creation of vdisk at EVA i have presented LUNS to the host. I have presented only 4 fcds but i don't know why 16 device files are created.
I am very much concern only on this due to that i thought i made a mistake.

[S.S.]

Can i remove the device file from which i created a vg to update the time stamp.
All alternate links time stamp is changed to today's date both in /dev/dsk and /dev/rdsk where as for the vg device file the time stamp is not updated on /dev/dsk.

The vg device file is showing the updated timestamp when i previous created the device file under the /dev/rdsk.