HPE Storage Users Group

Hello everyone.
After the node change I had several errors like this.
I fixed them all with "tunnel -f tp-1-sd-0.13". Only 3 I can not repair.

pdch ld:tp-1-sd-0.4 LD has 80 remote chunklets
pdch ld:tp-1-sd-0.13 LD has 86 remote chunklets
pdch ld:tp-1-sd-0.15 LD has 161 remote chunklets

Running tunneld -f xxxxxx gives me the error

2022-10-07 12:46:12 -03 Created task.
2022-10-07 12:46:12 -03 Updated Executing "tunnel tp-1-sd-0.13" as 0:14437
2022-10-07 12:46:13 -03 Updated ****
2022-10-07 12:46:13 -03 Updated **** tunnel started
2022-10-07 12:46:13 -03 Updated ****
2022-10-07 12:46:14 -03 Updated Error - unable to create new LD - error: Could not find enough available disk space.
2022-10-07 12:46:14 -03 Updated No tuning possible
2022-10-07 12:46:14 -03 Error
2022-10-07 12:46:14 -03 Error Task exited with status 1
2022-10-07 12:46:14 -03 Failed Could not complete task.

Any ideas?
as always Thanks.

Is your system very full and unbalanced?

You could try compactcpg as well.

MammaGutt wrote:Is your system very full and unbalanced?

You could try compactcpg as well.

Hello.
Thanks for your help.
It's all compacted. I have 22tb free.

sebastian7780_2 wrote:

MammaGutt wrote:Is your system very full and unbalanced?

You could try compactcpg as well.

Hello.
Thanks for your help.
It's all compacted. I have 22tb free.

Are you able to complete tuneld after running compactcpg?

If you have a balanced system and 22TB free capacity in the correct tier, you shouldn't have any issues running tuneld.

MammaGutt wrote:

sebastian7780_2 wrote:

MammaGutt wrote:Is your system very full and unbalanced?

You could try compactcpg as well.

Hello.
Thanks for your help.
It's all compacted. I have 22tb free.

Are you able to complete tuneld after running compactcpg?

If you have a balanced system and 22TB free capacity in the correct tier, you shouldn't have any issues running tuneld.

Mammagut,

I have run tunesys several times.
this is the checkhealth log -svc -detail

PD Cage:2 PDs FC/10K/900GB unbalanced. Primary path: 8 on Node:0, 6 on Node:1
pdch ld:tp-1-sd-0.4 LD has 80 remote chunklets
pdch ld:tp-1-sd-0.15 LD has 161 remote chunklets

I was able to repair only one LD after compacting and running "tunesys -chunkpct 1"

Showpd -path

---------Paths---------
Id CagePos Type -State- A B Order
0 2:8:0 FC normal 1:0:2\1:0:1 0:0:2\0:0:1 1/0
1 2:4:0 FC normal 1:0:2\1:0:1 0:0:2\0:0:1 0/1
3 2:6:0 FC normal 1:0:2\1:0:1 0:0:2\0:0:1 0/1
4 0:4:0 FC normal 1:0:1 0:0:1 1/0
5 0:5:0 FC normal 1:0:1 0:0:1 0/1
6 0:6:0 FC normal 1:0:1 0:0:1 1/0
7 0:7:0 FC normal 1:0:1 0:0:1 0/1
8 0:8:0 FC normal 1:0:1 0:0:1 1/0
9 0:9:0 FC normal 1:0:1 0:0:1 0/1
10 0:10:0 FC normal 1:0:1 0:0:1 1/0
11 0:11:0 FC normal 1:0:1 0:0:1 0/1
12 2:7:0 FC normal 1:0:2\1:0:1 0:0:2\0:0:1 1/0
13 2:9:0 FC normal 1:0:2\1:0:1 0:0:2\0:0:1 0/1
14 2:10:0 FC normal 1:0:2\1:0:1 0:0:2\0:0:1 1/0
15 2:11:0 FC normal 1:0:2\1:0:1 0:0:2\0:0:1 0/1
16 1:0:0 NL normal 1:0:1 0:0:1 1/0
17 1:1:0 NL normal 1:0:1 0:0:1 0/1
18 1:4:0 NL normal 1:0:1 0:0:1 0/1
19 1:5:0 NL normal 1:0:1 0:0:1 1/0
20 1:8:0 NL normal 1:0:1 0:0:1 1/0
21 1:9:0 NL normal 1:0:1 0:0:1 0/1
22 1:12:0 NL normal 1:0:1 0:0:1 0/1
23 1:13:0 NL normal 1:0:1 0:0:1 1/0
24 1:16:0 NL normal 1:0:1 0:0:1 1/0
25 1:17:0 NL normal 1:0:1 0:0:1 0/1
26 1:20:0 NL normal 1:0:1 0:0:1 0/1
27 1:21:0 NL normal 1:0:1 0:0:1 1/0
28 2:0:0 SSD normal 1:0:2 0:0:2 0/0
29 2:1:0 SSD normal 1:0:2 0:0:2 0/0
30 2:2:0 SSD normal 1:0:2 0:0:2 0/0
31 2:3:0 SSD normal 1:0:2 0:0:2 0/0
32 0:0:0 SSD normal 1:0:1\1:0:2 0:0:1\0:0:2 0/0
33 0:1:0 SSD normal 1:0:1\1:0:2 0:0:1\0:0:2 0/0
34 0:2:0 SSD normal 1:0:1\1:0:2 0:0:1\0:0:2 0/0
35 0:3:0 SSD normal 1:0:1\1:0:2 0:0:1\0:0:2 0/0
36 0:12:0 FC normal 1:0:1 0:0:1 1/0
37 0:13:0 FC normal 1:0:1 0:0:1 0/1
38 0:14:0 FC normal 1:0:1 0:0:1 1/0
39 0:15:0 FC normal 1:0:1 0:0:1 0/1
40 2:5:0 FC normal 1:0:2 0:0:2 0/1
41 0:17:0 FC normal 1:0:1 0:0:1 0/1
42 0:18:0 FC normal 1:0:1 0:0:1 1/0
43 0:19:0 FC normal 1:0:1 0:0:1 0/1
44 0:20:0 SSD normal 1:0:1 0:0:1 0/0
45 0:21:0 SSD normal 1:0:1 0:0:1 0/0
46 0:22:0 SSD normal 1:0:1 0:0:1 0/0
47 0:23:0 SSD normal 1:0:1 0:0:1 0/0
48 2:12:0 FC normal 1:0:2 0:0:2 1/0
49 2:13:0 SSD normal 1:0:2 0:0:2 0/0
50 2:14:0 SSD normal 1:0:2 0:0:2 0/0
51 2:15:0 FC normal 1:0:2 0:0:2 0/1
52 2:16:0 FC normal 1:0:2 0:0:2 1/0
53 2:17:0 FC normal 1:0:2 0:0:2 0/1
54 2:18:0 FC normal 1:0:2 0:0:2 1/0
55 2:19:0 FC normal 1:0:2 0:0:2 0/1
56 2:20:0 SSD normal 1:0:2 0:0:2 0/0
57 2:21:0 SSD normal 1:0:2 0:0:2 0/0
58 2:22:0 SSD normal 1:0:2 0:0:2 0/0
59 2:23:0 SSD normal 1:0:2 0:0:2 0/0
60 0:16:0 FC normal 1:0:1 0:0:1 1/0
-----------------------------------------------------
60 total

Do you think the balance error is related?
how can it be solved?
do you need any other log?
As I was saying, it all started after replacing a failed node

thanks for your help

what does

tuneld -f tp-1-sd-0.15

say?

Also, are you sure you got the cabling correct after node replacement?

Something looks funky.

MammaGutt wrote:what does

tuneld -f tp-1-sd-0.15

say?

Also, are you sure you got the cabling correct after node replacement?

Something looks funky.

Sure.
If I run "tunnel -f tp-1-sd-0.15" the work is never registered in the tasks. Instantly the mail arrives saying that it was not possible to execute... there is no space.

The task log is
2022-10-11 14:47:28 -03 Created task.
2022-10-11 14:47:28 -03 Updated Executing "tunnel tp-1-sd-0.15" as 0:14794
2022-10-11 14:47:29 -03 Updated ****
2022-10-11 14:47:29 -03 Updated **** tunnel started
2022-10-11 14:47:29 -03 Updated ****
2022-10-11 14:47:30 -03 Updated Error - unable to create new LD - error: Could not find enough available disk space.

2022-10-11 14:47:30 -03 Updated No tuning possible
2022-10-11 14:47:30 -03 Error
2022-10-11 14:47:30 -03 Error Task exited with status 1
2022-10-11 14:47:30 -03 Failed Could not complete task.

the cables are ok.

3par_Ducasse cli% checkhealth cabling
Checking wiring
The following components are healthy: cabling

3par_Ducasse cli% showspace
--Estimated(MB)---
RawFree UsableFree
5931008 2965504
3par_Ducasse cli%

Thanks.

Okay....

tp-1..... means this is a LD for CPG #1

Issue "showcpg" and find the CPG with ID 1.

Then do "showspace -cpg <cpgname>"

Then do "showpd -c -p -devtype <SSD , FC or NL depending on CPG>". Look at how much free space you have for that tier.

You could do "showpd" ... The path showing "*" indicates which node owns a PD. Verify that this is equal on both nodes.

You can do "showld" to see the size of the two LDs. A normal LD is usually 256GiB. If you have LDs that are bigger (SizeMB) than their usage (UsedMB), you should be able to extract that capacity by doing "compactcpg -f <cpgname>" .... Remember, first digit in a LD name is the CPG ID.... And you could expect more than a few GiB before bothering to do compactcpg on a CPG.

MammaGutt wrote:Okay....

tp-1..... means this is a LD for CPG #1

Issue "showcpg" and find the CPG with ID 1.

Then do "showspace -cpg <cpgname>"

Then do "showpd -c -p -devtype <SSD , FC or NL depending on CPG>". Look at how much free space you have for that tier.

You could do "showpd" ... The path showing "*" indicates which node owns a PD. Verify that this is equal on both nodes.

You can do "showld" to see the size of the two LDs. A normal LD is usually 256GiB. If you have LDs that are bigger (SizeMB) than their usage (UsedMB), you should be able to extract that capacity by doing "compactcpg -f <cpgname>" .... Remember, first digit in a LD name is the CPG ID.... And you could expect more than a few GiB before bothering to do compactcpg on a CPG.

Mammagut.
I tell you...
after running several AO, compact... and so many times little by little I was repairing the LD.

Thank you again.

sebastian7780_2 wrote:

MammaGutt wrote:Okay....

tp-1..... means this is a LD for CPG #1

Issue "showcpg" and find the CPG with ID 1.

Then do "showspace -cpg <cpgname>"

Then do "showpd -c -p -devtype <SSD , FC or NL depending on CPG>". Look at how much free space you have for that tier.

You could do "showpd" ... The path showing "*" indicates which node owns a PD. Verify that this is equal on both nodes.

You can do "showld" to see the size of the two LDs. A normal LD is usually 256GiB. If you have LDs that are bigger (SizeMB) than their usage (UsedMB), you should be able to extract that capacity by doing "compactcpg -f <cpgname>" .... Remember, first digit in a LD name is the CPG ID.... And you could expect more than a few GiB before bothering to do compactcpg on a CPG.

Mammagut.
I tell you...
after running several AO, compact... and so many times little by little I was repairing the LD.

Thank you again.

Ahhh, you're running AO.... AO, compactcpg and tuneld/tunesys are three tasks that are fighting against eachother. If two of them run at the same time you're 99% sure that at least one will fail (or not fully complete). Next time, disable AO before you start and re-enable when you're done Then you'll be able to complete on the first attempt