Ivo Fortunov Posted June 2, 2020 Share #1 Posted June 2, 2020 Hello, I'm trying to create a high availability datastore on two ESXi hosts on two physical machines. I'm trying with both loaders v1.03b (ds3615xs and ds3617xs). The High Availability setup completes successfully and at the end restarts the passive node. After the passive node starts Synology High Availability says "Unable to detect passive server". On the Network->Hosts tab it says that both interfaces on both hosts are connected. Ping on LAN1 on both hosts is OK. I have tested the setup on VMWare Workstation (both on 2 direct VMs and 2 virtualized ESXi hosts with 2 VWs inside) and everything works fine. Any ideas how to diagnose the problem? Quote Link to comment Share on other sites More sharing options...
orcodox Posted June 14, 2020 Share #2 Posted June 14, 2020 Hi, I have the exact same issue. What's weird about it, is that the actual replication is happening. Just the UI is not representing that. Are you saying this is not the case on 2 virtualised esxi hosts with the exact same setup? Also, since it has been a while since u posted. Have you figured it out? Quote Link to comment Share on other sites More sharing options...
Ivo Fortunov Posted June 26, 2020 Author Share #3 Posted June 26, 2020 When I install the ESXi hosts on VMWare Workstation everything works fine. It is not working on physical servers. Quote Link to comment Share on other sites More sharing options...
Aigor Posted July 6, 2020 Share #4 Posted July 6, 2020 You need virtual serial port for both VM, 2 network cards, one for regular network, other for heartbeat. Heartbeat network should be visible only for both machine. Quote Link to comment Share on other sites More sharing options...
Ivo Fortunov Posted July 6, 2020 Author Share #5 Posted July 6, 2020 Thanks Aigor, but that is exactly my setup. Quote Link to comment Share on other sites More sharing options...
aw5186 Posted March 23, 2021 Share #6 Posted March 23, 2021 Hi, I have the exact same issue. What's weird about it, is that the actual replication is happening. Just the UI is not representing that. esxi 6.5-6.7u3 all test host A dell r720xd e5-2680 v2 host B hp dl380g9 e5-2680 v4 host C hp dl80g9 e5 2604 v3 ds3615xs,ds3617xs, 1.03b load dsm 6.2.3 hostA --- hostB success; hostB ---host A fail host B --- host C or host C ----hostB all success hostA --- host C success; host C---hostA fail Try to enable the host B VMware EVC mode as Intel "Ivy bridge" generation, which is consistent with the architecture of host a, hostA --- hostB ;hostB ---host A all fail。 cluster.log corosync.log ha.daemon.log ha.daemon.record lsyncd_system_sync.log Quote Link to comment Share on other sites More sharing options...
aw5186 Posted March 23, 2021 Share #7 Posted March 23, 2021 (edited) Through continuous testing, we finally found out the reason: It is caused by the CPU architecture, because the Intel "Ivy bridge" generation architecture of e52680 V2 is too low, resulting in abnormal operation of the syntax high availability process The preliminary judgment is that the syntax high availability uses the new AES encryption algorithm instruction set, but Ivy bridge architecture does not have this instruction set,but, Looking up the data, it is found that e52680v2 supports AES Ni advanced instruction set, so it is not the cause of this instruction set. Finally, it is determined that the operation failure of syntax high availability is caused by ivy bridge architecture, After I set the VMware EVC mode of DSM VM of HP DL 380g9 to Intel "Ivy bridge" generation, Host a VM DSM and host B VM DSM failed to create synchronization high availability, The system internal service [syntax cron daemon] failed to start After canceling VMware EVC Hosta -- host B succeeds, hostb -- Hosta fails; Cancel VMware EVC mode successfully hostA --- host B Switch active server to passive server, hostb -- host a failed Edited March 23, 2021 by aw5186 1 Quote Link to comment Share on other sites More sharing options...
aw5186 Posted March 24, 2021 Share #8 Posted March 24, 2021 (edited) 22 hours ago, aw5186 said: Through continuous testing, we finally found out the reason: It is caused by the CPU architecture, because the Intel "Ivy bridge" generation architecture of e52680 V2 is too low, resulting in abnormal operation of the syntax high availability process The preliminary judgment is that the syntax high availability uses the new AES encryption algorithm instruction set, but Ivy bridge architecture does not have this instruction set,but, Looking up the data, it is found that e52680v2 supports AES Ni advanced instruction set, so it is not the cause of this instruction set. Finally, it is determined that the operation failure of syntax high availability is caused by ivy bridge architecture, After I set the VMware EVC mode of DSM VM of HP DL 380g9 to Intel "Ivy bridge" generation, Host a VM DSM and host B VM DSM failed to create synchronization high availability, The system internal service [syntax cron daemon] failed to start After canceling VMware EVC Hosta -- host B succeeds, hostb -- Hosta fails; Cancel VMware EVC mode successfully hostA --- host B Switch active server to passive server, hostb -- host a failed Edited March 24, 2021 by aw5186 Important Note: After testing, although the HA status is shown as failure, all the Settings and data can be synchronized successfully, and automatic switching can be achieved, but manual switching is not possible. Quote Link to comment Share on other sites More sharing options...
aw5186 Posted March 24, 2021 Share #9 Posted March 24, 2021 Important Note: After testing, although the HA status is shown as failure, all the Settings and data can be synchronized successfully, and automatic switching can be achieved, but manual switching is not possible. Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.