New answers tagged

0

So, I was found a solution. Open the file /etc/iscsi/nodes/iqn.2000-node-name/10.0.3.89,3260,1/default Insert next rows: node.session.auth.username = user_defined_in_target_setting node.session.auth.password = your_password Login to target from initiator should be successful.


3

Ok.. it's been a while... I was hoping some has done some testing and can help me in this matter, but after last few days, I've been testing some configurations and benchmark them, trying to find and compensate myself. This may not be perfect config, but this is comfortable enough for me. I go with Windows 2012 HyperV + 2008 R2. StarWind running at windows ...


0

I have zero experience with opensuse, but this is what we do with our network appliance targets in centos. Steps Open the /etc/iscsi/iscsid.conf file with a text editor. Enable CHAP authentication by setting node.session.auth.authmethod to CHAP, the default is None. node.session.auth.authmethod = CHAP Provide a CHAP user name and password for the ...


1

Infiniband would be 40 Gigabits per second. That's 5000 Megabytes per second. Your drives combined output (300MB/sec * 16) is 4800 Megabytes per second (38.4Gb/s). Sounds like you are maxing out your Infiniband. I don't believe the common port concern is your issue. Unless I've misinterpreted your configuration?


2

Qemu takes VMs into pause on io errors by design, so that you will avoid corrupt disks. All you have to do is restore the iscsi connections and run/unpause the VMs


4

Modern OS and filesystem are much more resilient against corruption, and the same can be said for MySQL (especially when used with InnoDB tables). Basically, anything written to disk with sync/fsync semantic should be crash-safe, as the writes will not return until the data are on stable storage. Moreover, InnoDB uses internal journaling to be sure no ...


1

Edit: Even after the switch got power again and the storage was available, the VM's were still in an unresponsive crashed state. They did not react to any input whatsoever forcing me to do a hard-reset one by one, relying on the filesystem journaling to repair the damage. I was lucky it worked and nothing borked. As far as I understand, I got lucky there as ...


3

Network filesystems and databases can be a problematic combo, and at least InnoDB does not like NFS. From the MySQL documentation If reliability is a consideration for your data, do not configure InnoDB to use data files or log files on NFS volumes. Potential problems vary according to OS and version of NFS, and include such issues as lack of ...


3

StarWind vSAN runs under VMware just fine. You have VMware's own vSAN there also. Shouldn't be an issue to live migrate your VMs.


0

this is an issue with any Microsoft failover cluster using Server 2012 or below using either disk majority or node and disk majority. The Disk witness is a point of failure for the cluster. Even if all nodes are up, if the disk witness fails (we loose access to it) the cluster goes down. so what to do? 1) Make sure your disk witness is highly available so ...



Top 50 recent answers are included