Unsolved
This post is more than 5 years old
44 Posts
0
4660
EMC RecoverPoint SRA does not discover any datastores on replicated devices
Hello,
I experienced some unexpected results with my new setup of RPA and SRM.
My RecoverPoint cluster is v4.1.P1(d.219).
SRM is v5.5.1 with EMC RecoverPoint SRA v 2.2.0.
Beside that there is EMC VSI for vSphere installed v.5.6.1.2.
I created two new CGs on RPA with 4 Replication Sets each. Group Policy for my CGs is set as external application, SRM in my case.
CG is Active and Synchronized.
After synchronization finished the CG status still shows me "Storage: No Access" at both sides, but for remote copy image I see:Distributing Pre-replication image fast forward. I am not sure the meaning of this....?
The other, old CGs reports the storage access status as "Direct Access" and local part of replication set reports it status as "Writable and Splitting".
This is what I expect for my new CGs, but I do not know how to fix their state.
Moreover, "Storage: No Access" status is not really true, as local LUNs are still mounted on vsphere hosts.
Beside that I noticed a problem at the SRA level.
Replicated Devices are visible in SRM Array Managers with the right replication direction arrow, but SRA does not discover any datastores.
This prevents me to use that devices with new protection group.
I would appreciate for any help or solving suggestion.
Regards
--
Darek
etaljic81
1K Posts
0
October 15th, 2014 04:00
Can you verify that the LUNs are attached to the splitter in RP?
darekwj
44 Posts
0
October 15th, 2014 05:00
Hi,
I have VNX arrays at the backend.
I assume You mean if the LUNs are masked to RPA Storage Groups - yes they are.
Affected LUNs have ALU numbers 171-178.
The other that behave properly are 151 and 152.
RPA Storage groups:
HLU/ALU Pairs:
HLU Number ALU Number
---------- ----------
0 16
17 17
18 18
152 152
151 151
19 19
20 20
21 21
22 22
1 171
2 177
3 173
4 175
5 176
6 178
7 172
8 174
Hosts Storagegroup ex:
HLU/ALU Pairs:
HLU Number ALU Number
---------- ----------
0 4
134 134
135 135
139 139
142 142
143 143
1 144
2 145
3 146
4 147
148 148
160 160
161 161
162 162
163 163
164 164
170 170
171 171
173 173
172 172
174 174
175 175
177 177
176 176
178 178
Shareable: YES
In Unisphere for RecoverPoint I see that volumes are attached to splitter, but their Access is reported as "No Access"
Can different HLU on RPA and on host cause the problem?
etaljic81
1K Posts
0
October 15th, 2014 06:00
No, different HLUs between the RPAs and the hosts don't matter.
darekwj
44 Posts
0
October 15th, 2014 07:00
Hi,
The Image Status changed during night from: "Distributing Pre-Replication image Fast Forward" to normal "Distributing" state with indication to the last snapshot date. But I still see "Storage No Access" both at Production and Remote Copy side.
get_group_settings used with my new CGs shows :Fail all production: YES, For other working CGs this attribute is set an NO.
darekwj
44 Posts
0
October 20th, 2014 01:00
I removed my CGs and recreated them from scratch. The new CGs initiated properly and by this time SRA has detected datastores correctly. I configured the new Protection Groups and Recovery Plans on SRM. All operations finished successfully.
The situation which I experienced at the previous trial, it must be some unusual anomaly...
odurasler757
25 Posts
0
October 24th, 2014 11:00
hi darekwj, did you have multiple copies of a CG to two or more sites? I ask because, i'm having an issue where i have multiple sites and it seems SRA is unable to discover my CGs. However, CGs with only 1 copy have no issues.
I'll probably start a new thread if it's not a similar issue.
darekwj
44 Posts
0
October 27th, 2014 00:00
HI,
My CGs contain four replications sets, but I think in my case it makes no difference.
I asked support to investigate and They found “Fail all Production” flag on my CG. From what I learned It is related to failed or unsuccessfull SRM or vsphere operation , ex. when SRM recovery is either aborted or failed before completion this flag appears at the output of get_group_settings command.
It is possible to use CLI to solve the problem: resume_writes_after_srm command resumes writes to production after an aborted or unsuccessful failover attempt.
forshr
1.1K Posts
0
October 27th, 2014 03:00
Yes, this will reset the CG at the Prod site by setting a state of R/W or direct access for the prod volumes.
etaljic81
1K Posts
0
October 27th, 2014 10:00
Glad you got it fixed. Definitely something to keep in mind.
smitty2k1
1 Message
0
June 5th, 2015 08:00
did you ever resolve that issue ? Experiencing the same problem