| Asset ID: |
1-72-2337026.1 |
| Update Date: | 2017-12-08 |
| Keywords: | |
Solution Type
Problem Resolution Sure
Solution
2337026.1
:
Oracle ZFS Storage Appliance: Replication Fails with Connection Timeouts
| Related Items |
- Sun ZFS Storage 7420
- Oracle ZFS Storage ZS5-2
- Oracle ZFS Storage ZS3-2
- Sun Storage 7110 Unified Storage System
- Oracle ZFS Storage ZS4-4
- Sun Storage 7210 Unified Storage System
- Oracle ZFS Storage ZS5-4
- Sun Storage 7410 Unified Storage System
- Oracle ZFS Storage ZS3-4
- Sun ZFS Storage 7120
- Sun Storage 7310 Unified Storage System
- Sun ZFS Storage 7320
|
| Related Categories |
- PLA-Support>Sun Systems>DISK>ZFS Storage>SN-DK: ZS
|
In this Document
Created from <SR 3-16347245241>
Applies to:
Oracle ZFS Storage ZS5-4 - Version All Versions and later
Oracle ZFS Storage ZS5-2 - Version All Versions and later
Oracle ZFS Storage ZS4-4 - Version All Versions and later
Oracle ZFS Storage ZS3-4 - Version All Versions and later
Oracle ZFS Storage ZS3-2 - Version All Versions and later
7000 Appliance OS (Fishworks)
Symptoms
Configuration notes:
- Using two VNICs which were then paired into active/passive IPMP interface
- VLANs were in use and the VNICs were both assigned the same VLAN
- MTU was 9000
- Target and source were configured similarly
Observations
- The ports 216 and 217 ( in fact there was no firewall between ) were open
- We could see that when a replication was started it successfully bound to the target
- snoop indicated packet flow initially , but stopped totally after about 20 seconds
- Source eventually failed with: errmsg = stage 'client_connect' failed: ssl connection timed out
- Target failed with: error = ak_stream_buffer_reader() failed: SSL read i/o system call
Replication had been working, then some network changes had been made.
Now we see that replication starts, source connects to target TCP/IP ports, but then times out without sending any data
On the source:
Tue Sep 12 17:35:08 2017
nvlist version: 0
time = 0x59b81acc
hrtime = 0x1dc83785989110
action = (embedded nvlist)
nvlist version: 0
target_label = npaptzfs1h1-repl.servicemngt.ageas.intra
target_uuid = d7097a1f-792c-c0af-eef4-c92808521d1e
uuid = ffd30624-4404-69b6-9457-c74c7ba38b0a
state = sending
dataset = h1-raidz2/local/BACKUPSET1_A00PR
(end action)
event = update done
result = failure
errmsg = stage 'client_connect' failed: ssl connection timed out
remote_status = ok
On the target:
Tue Sep 12 17:40:48 2017
nvlist version: 0
time = 0x59b81c20
hrtime = 0x14f8d6381ea181
pkg = (embedded nvlist)
nvlist version: 0
source_asn = a61e55e2-26da-e85d-9bf7-c0ecfc18c8e0
source_name = praptzfs1h2
uuid = ffd30624-4404-69b6-9457-c74c7ba38b0a
state = receiving
(end pkg)
event = recv_done
result = failed
error = ak_stream_buffer_reader() failed: SSL read i/o system call
error (Connection timed out)
Changes
Network changes had been made.
Cause
The ZFS-SA datalinks were configured for Jumbo Packets (MTU = 9000).
When using the non default packet size (MTU), each interconnecting component (ie. switches) must also support the MTU in use.
Solution
Resolved this by changing the MTU from 9000 to the default 1500 on the datalinks on the source and target ZFSSA to match the interconnecting network equipment.
Attachments
This solution has no attachment
