Snapshot can't be created - Expired can't be deleted

[Omaroo7]

Hello
I have 3par8200 I am running a schedule to take snapshot by system .
I found some LUNs fail creating a new snapshot , also new snapshots can't be created .
Error shown as screen shot. Show That tree is being Cleaned ,
From few Months and the tree is being cleaned still exist.
I can't create snapshot manually or by schedule I can't delete expired snapshots manually or by schedule.
Cleaning the tree is still show for months

What about a command that force expired snapshot to be cleared??
What should I do ?

[Omaroo7]

Can Any one HELP ?

[sbhat09]

Hello @Omaroo7,

For the volumes which the snapshots are failing, Please change the snap CPG and try to take snapshots. If you continue to get the error, please click on the 'details' of the error message above and share further details.

Regards,
Srinivas Bhat

If you feel this was helpful please click the KUDOS! thumb below!
Note: All of my comments are my own and are not any official representation of HPE.

[Omaroo7]

I tryed to change the copy CPG But I get this error , I tryied many other cpg to be as copy cpg I get the same error even , Even the CPG have free space

When I run VVshow I get this  "The last snapshot show not_started " as you can see in below screenshot

[sbhat09]

Hello @Omaroo7,

Can you please click on 'Details' there and share the screenshot of further details?

Regards,
Srinivas Bhat

If you feel this was helpful please click the KUDOS! thumb below!
Note: All of my comments are my own and are not any official representation of HPE.

[Omaroo7]

[sbhat09]

Hello @Omaroo7,

This looks strange.! I think it requires some tests to understand the cause behind this behaviour.

Please check if there are any orphan snapshots or clones. If yes, please remove them.

To check if the issue is with the copy CPG, please change the copy CPG to any other CPG where there is enough space and check the snapshot behaviour. (I think you already tested this)

To check if the issue is with the volumes, please schedule the snapshots with some other volumes keeping the same copy CPG where the snapshots are failing.

To check if there are any issues with scheduling, please try to take some one-time snapshots keeping difeerent copy CPGs.

Hope these test results must bring more clarity about the issue.

Regards,
Srinivas Bhat

If you feel this was helpful please click the KUDOS! thumb below!
Note: All of my comments are my own and are not any official representation of HPE.

[Omaroo7]

As I said
I can't create a one time snapshot U get error
"Virtual volume Tree is being cleaned try later"

I can't change the copy cpg also I get the error "use tunevv to allocate snapshot space from a different CPG"
Even Thier is free space and I tried two different CPG I get the same error message ?!!

I trying to remove the snapshots but get error can't be deleted check object to be removed from the cash

The snapshot show status is failed

I tried to change the schedule but also the new schedule snapshot failed

any way to force remove volume snapshots????

[sbhat09]

Hello @Omaroo7,

To force remove all the snapshots, you can use the command 'removevv -f -snaponly'.

Regards,
Srinivas Bhat

If you feel this was helpful please click the KUDOS! thumb below!
Note: All of my comments are my own and are not any official representation of HPE.

[Omaroo7]

 The snapshot show fail status as follow ,

I tried to remove the snapshot using  -f but the snapshot still exist

Status of the snapshot is failed befor running remove command

[Omaroo7]

what about using -stale instead of -snaponly for the command :    "removevv -stale vvname "

Where vvname is the name of the failed snapshot  ,

would it work? , have any impact on base LUN ?

what about using -expired instead of -snaponly for the command :    "removevv -expired vvname "

Where vvname is the name of the failed snapshot  ,

would it work? , have any impact on base LUN ?

[sbhat09]

Hello @Omaroo7,

The commands 'removevv -stale' or 'removevv -expired' will not impact the base VVs. But before that run the command 'showvv' and check the columns 'type' and 'Detailed_State'. Those VVs 'base' and 'normal' will not be impacted by those commands.

Regards,
Srinivas Bhat

If you feel this was helpful please click the KUDOS! thumb below!
Note: All of my comments are my own and are not any official representation of HPE.

[sbhat09]

Hello @Omaroo7 

Were you able to remove the expired snapshots successfully?

Regards,
Srinivas Bhat

If you feel this was helpful please click the KUDOS! thumb below!
Note: All of my comments are my own and are not any official representation of HPE.

[Omaroo7]

No not yet
Could this issue related to some bug in the current 3par OS version?

Could any one get access to the Linux os level for 3 par and check on the Linux OS level or this can be done by HPE support only ?

[sbhat09]

Hello @Omaroo7,

Ok. Further troubleshooting must be done by HPE support higher tier teams.

Regards,
Srinivas Bhat

If you feel this was helpful please click the KUDOS! thumb below!
Note: All of my comments are my own and are not any official representation of HPE.

[Omaroo7]

what about system tasks and task manager ? do we have some tasks depend on other task?

I mean maybe creating snapshots and deleting expired snapshots task is depend on  a system task

What is the available system task that are core for the system and other tasks have dependant on it?

[sbhat09]

Hello @Omaroo7,

There are no such dependency of any tasks for the creation or deletion of snapshots. Only dependency is that those snapshots must not be exported to any hosts or they must not be in the range of the protection period.

Regarding tasks, they are categorised based on priority high, medium and low. System prioritizes higher priority tasks in case of multiple tasks are queued up.

Regards,
Srinivas Bhat

If you feel this was helpful please click the KUDOS! thumb below!
Note: All of my comments are my own and are not any official representation of HPE.

[Omaroo7]

I run the command checkvv on one of base volumes with failed snapshots and I run show for the task and get this output

******cli% showtask -d 8750
Id Type Name Status Phase Step -------StartTime------- ------FinishTime------- -Priority- -User--
8750 check_vv checkvv done --- --- 2022-10-21 19:14:24 EET 2022-10-21 19:14:31 EET n/a *******

Detailed status:
2022-10-21 19:14:24 EET Created task.
2022-10-21 19:14:24 EET Started checkvv process for VVs. checkvv options: -n
2022-10-21 19:14:31 EET Updated VV 'VOL-SSD.220928000102' v_tptype 0 should not have v_elroot vt_type 0.
2022-10-21 19:14:31 EET Updated checkvv process for VV VolSSD
2022-10-21 19:14:31 EET Finished checkvv process

[Omaroo7]

Any help please

[sbhat09]

Hello @Omaroo7 

I think you have run the command 'checkvv -n VV_name'. This command only checks the base volume for any errors. But (the option '-n') doesn't modify if the errors are found.
So, this doesn't help fix anything in this concern.
If you are intended to check the VV and also fix if something was wrong, you can run 'checkvv -n VV_name'

Regards,
Srinivas Bhat

If you feel this was helpful please click the KUDOS! thumb below!
Note: All of my comments are my own and are not any official representation of HPE.