Re: Issue with GRE Tunnel between a VM and outside server
Hi Wei, I highly appreciate your help. Your proposed solution worked for me. Thanks. On Sat, 20 Jan 2024, 16:36 Wei ZHOU, wrote: > Have you retried after adding security group rule with protocol number = > 47? > > -Wei > > > 在 2024年1月20日星期六,Fariborz Navidan 写道: > > > Hi Daan, > > > > We still couldn't sort out this issue with our client VM. We are still > > waiting for the community to direct us toward finding a solution. > > > > Regards. > > > > On Fri, 19 Jan 2024, 17:31 Daan Hoogland, > wrote: > > > > > Friborz, any progress? > > > not a gre expert but glad to see you get on with your problem. > > > > > > On Sat, Jan 6, 2024 at 10:39 PM Fariborz Navidan < > mdvlinqu...@gmail.com> > > > wrote: > > > > > > > > Hi Dear Experts, > > > > > > > > We are running Cs 4.15.0.0 with 2 KVM hosts having security groups > > > enabled > > > > zone. We have a VM which a GRE tunnel has been setup between it and a > > > > server outside our network. Both hosts had been rebooted a few days > ago > > > due > > > > to power interruption. Before the reboot happens, the GRE tunnel was > > > > working properly on the mentioned VM. However after the reboot, GRE > > > tunnel > > > > can be established but machines cannot reach each other via the > > tunnel's > > > > private IP address. All ports and protocols are already added to > > ingress > > > > rule set of security group which VM belongs to. > > > > > > > > Below is output of "ip a" and "ip r" commands on the VM running on > our > > CS > > > > infrastructure. > > > > > > > > root@cdn-fr-1-kajgana-net:~# ip a > > > > 1: lo: mtu 65536 qdisc noqueue state UNKNOWN > > group > > > > default qlen 1000 > > > > link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 > > > > inet 127.0.0.1/8 scope host lo > > > >valid_lft forever preferred_lft forever > > > > inet6 ::1/128 scope host > > > >valid_lft forever preferred_lft forever > > > > 2: ens3: mtu 1500 qdisc pfifo_fast > > > state > > > > UP group default qlen 1000 > > > > link/ether 1e:00:85:00:02:4d brd ff:ff:ff:ff:ff:ff > > > > inet 164.132.223.34/28 brd 164.132.223.47 scope global ens3 > > > >valid_lft forever preferred_lft forever > > > > inet6 fe80::1c00:85ff:fe00:24d/64 scope link > > > >valid_lft forever preferred_lft forever > > > > 3: gre0@NONE: mtu 1476 qdisc noop state DOWN group default > > qlen > > > 1000 > > > > link/gre 0.0.0.0 brd 0.0.0.0 > > > > 4: gretap0@NONE: mtu 1462 qdisc noop state > DOWN > > > group > > > > default qlen 1000 > > > > link/ether 00:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff > > > > 5: erspan0@NONE: mtu 1450 qdisc noop state > DOWN > > > group > > > > default qlen 1000 > > > > link/ether 00:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff > > > > 6: gre1@NONE: mtu 1476 qdisc noqueue > > > state > > > > UNKNOWN group default qlen 1000 > > > > link/gre 164.132.223.34 peer 89.205.123.34 > > > > inet 192.168.169.1/30 scope global gre1 > > > >valid_lft forever preferred_lft forever > > > > inet6 fe80::200:5efe:a484:df22/64 scope link > > > >valid_lft forever preferred_lft forever > > > > > > > > root@cdn-fr-1-kajgana-net:~# ip r > > > > default via 164.132.223.46 dev ens3 > > > > 164.132.223.32/28 dev ens3 proto kernel scope link src > 164.132.223.34 > > > > 192.168.169.0/30 dev gre1 proto kernel scope link src 192.168.169.1 > > > > > > > > IP address of tunnel's other endpoint is 192.168.169.2 which is > > > unreachable > > > > from the VM. It looks like GRE tunnel has been established but > traffic > > > > cannot be p[assed through. > > > > > > > > Is there something we need to do with iptables rules on the hosts to > > > allow > > > > GRE traffic or is there anything else we can do to address this > issue? > > > > > > > > Thanks in advance. > > > > Regards. > > > > > > > > > > > > -- > > > Daan > > > > > >
Re: Issue with GRE Tunnel between a VM and outside server
Have you retried after adding security group rule with protocol number = 47? -Wei 在 2024年1月20日星期六,Fariborz Navidan 写道: > Hi Daan, > > We still couldn't sort out this issue with our client VM. We are still > waiting for the community to direct us toward finding a solution. > > Regards. > > On Fri, 19 Jan 2024, 17:31 Daan Hoogland, wrote: > > > Friborz, any progress? > > not a gre expert but glad to see you get on with your problem. > > > > On Sat, Jan 6, 2024 at 10:39 PM Fariborz Navidan > > wrote: > > > > > > Hi Dear Experts, > > > > > > We are running Cs 4.15.0.0 with 2 KVM hosts having security groups > > enabled > > > zone. We have a VM which a GRE tunnel has been setup between it and a > > > server outside our network. Both hosts had been rebooted a few days ago > > due > > > to power interruption. Before the reboot happens, the GRE tunnel was > > > working properly on the mentioned VM. However after the reboot, GRE > > tunnel > > > can be established but machines cannot reach each other via the > tunnel's > > > private IP address. All ports and protocols are already added to > ingress > > > rule set of security group which VM belongs to. > > > > > > Below is output of "ip a" and "ip r" commands on the VM running on our > CS > > > infrastructure. > > > > > > root@cdn-fr-1-kajgana-net:~# ip a > > > 1: lo: mtu 65536 qdisc noqueue state UNKNOWN > group > > > default qlen 1000 > > > link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 > > > inet 127.0.0.1/8 scope host lo > > >valid_lft forever preferred_lft forever > > > inet6 ::1/128 scope host > > >valid_lft forever preferred_lft forever > > > 2: ens3: mtu 1500 qdisc pfifo_fast > > state > > > UP group default qlen 1000 > > > link/ether 1e:00:85:00:02:4d brd ff:ff:ff:ff:ff:ff > > > inet 164.132.223.34/28 brd 164.132.223.47 scope global ens3 > > >valid_lft forever preferred_lft forever > > > inet6 fe80::1c00:85ff:fe00:24d/64 scope link > > >valid_lft forever preferred_lft forever > > > 3: gre0@NONE: mtu 1476 qdisc noop state DOWN group default > qlen > > 1000 > > > link/gre 0.0.0.0 brd 0.0.0.0 > > > 4: gretap0@NONE: mtu 1462 qdisc noop state DOWN > > group > > > default qlen 1000 > > > link/ether 00:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff > > > 5: erspan0@NONE: mtu 1450 qdisc noop state DOWN > > group > > > default qlen 1000 > > > link/ether 00:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff > > > 6: gre1@NONE: mtu 1476 qdisc noqueue > > state > > > UNKNOWN group default qlen 1000 > > > link/gre 164.132.223.34 peer 89.205.123.34 > > > inet 192.168.169.1/30 scope global gre1 > > >valid_lft forever preferred_lft forever > > > inet6 fe80::200:5efe:a484:df22/64 scope link > > >valid_lft forever preferred_lft forever > > > > > > root@cdn-fr-1-kajgana-net:~# ip r > > > default via 164.132.223.46 dev ens3 > > > 164.132.223.32/28 dev ens3 proto kernel scope link src 164.132.223.34 > > > 192.168.169.0/30 dev gre1 proto kernel scope link src 192.168.169.1 > > > > > > IP address of tunnel's other endpoint is 192.168.169.2 which is > > unreachable > > > from the VM. It looks like GRE tunnel has been established but traffic > > > cannot be p[assed through. > > > > > > Is there something we need to do with iptables rules on the hosts to > > allow > > > GRE traffic or is there anything else we can do to address this issue? > > > > > > Thanks in advance. > > > Regards. > > > > > > > > -- > > Daan > > >
Re: Issue with GRE Tunnel between a VM and outside server
Hi Daan, We still couldn't sort out this issue with our client VM. We are still waiting for the community to direct us toward finding a solution. Regards. On Fri, 19 Jan 2024, 17:31 Daan Hoogland, wrote: > Friborz, any progress? > not a gre expert but glad to see you get on with your problem. > > On Sat, Jan 6, 2024 at 10:39 PM Fariborz Navidan > wrote: > > > > Hi Dear Experts, > > > > We are running Cs 4.15.0.0 with 2 KVM hosts having security groups > enabled > > zone. We have a VM which a GRE tunnel has been setup between it and a > > server outside our network. Both hosts had been rebooted a few days ago > due > > to power interruption. Before the reboot happens, the GRE tunnel was > > working properly on the mentioned VM. However after the reboot, GRE > tunnel > > can be established but machines cannot reach each other via the tunnel's > > private IP address. All ports and protocols are already added to ingress > > rule set of security group which VM belongs to. > > > > Below is output of "ip a" and "ip r" commands on the VM running on our CS > > infrastructure. > > > > root@cdn-fr-1-kajgana-net:~# ip a > > 1: lo: mtu 65536 qdisc noqueue state UNKNOWN group > > default qlen 1000 > > link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 > > inet 127.0.0.1/8 scope host lo > >valid_lft forever preferred_lft forever > > inet6 ::1/128 scope host > >valid_lft forever preferred_lft forever > > 2: ens3: mtu 1500 qdisc pfifo_fast > state > > UP group default qlen 1000 > > link/ether 1e:00:85:00:02:4d brd ff:ff:ff:ff:ff:ff > > inet 164.132.223.34/28 brd 164.132.223.47 scope global ens3 > >valid_lft forever preferred_lft forever > > inet6 fe80::1c00:85ff:fe00:24d/64 scope link > >valid_lft forever preferred_lft forever > > 3: gre0@NONE: mtu 1476 qdisc noop state DOWN group default qlen > 1000 > > link/gre 0.0.0.0 brd 0.0.0.0 > > 4: gretap0@NONE: mtu 1462 qdisc noop state DOWN > group > > default qlen 1000 > > link/ether 00:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff > > 5: erspan0@NONE: mtu 1450 qdisc noop state DOWN > group > > default qlen 1000 > > link/ether 00:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff > > 6: gre1@NONE: mtu 1476 qdisc noqueue > state > > UNKNOWN group default qlen 1000 > > link/gre 164.132.223.34 peer 89.205.123.34 > > inet 192.168.169.1/30 scope global gre1 > >valid_lft forever preferred_lft forever > > inet6 fe80::200:5efe:a484:df22/64 scope link > >valid_lft forever preferred_lft forever > > > > root@cdn-fr-1-kajgana-net:~# ip r > > default via 164.132.223.46 dev ens3 > > 164.132.223.32/28 dev ens3 proto kernel scope link src 164.132.223.34 > > 192.168.169.0/30 dev gre1 proto kernel scope link src 192.168.169.1 > > > > IP address of tunnel's other endpoint is 192.168.169.2 which is > unreachable > > from the VM. It looks like GRE tunnel has been established but traffic > > cannot be p[assed through. > > > > Is there something we need to do with iptables rules on the hosts to > allow > > GRE traffic or is there anything else we can do to address this issue? > > > > Thanks in advance. > > Regards. > > > > -- > Daan >
Re: Issues migrating primary storage
I’m trying to put my NFS primary storage in to maintenance mode, which I believe is supposed to migrate all of its storage, correct? The problem is I don’t know how to get a status on this job? I can’t really tell if it’s working. Management server doesn’t really have anything in the logs…. I don’t see any new images or images growing on the Ceph side. So I just don’t know if it’s working or how far along the migration is. -jeremy > On Friday, Jan 19, 2024 at 12:34 AM, Jeremy Hansen (mailto:jer...@skidrow.la)> wrote: > I’m still having issues. Is it unreasonable to migrate 1TB images over a 10G > network? Any other ideas of things to try would be much appreciated. > > -jeremy > > > > > On Wednesday, Jan 17, 2024 at 12:49 PM, Jeremy Hansen > (mailto:jer...@skidrow.la)> wrote: > > Extending these timeouts in the “wait” configs seems to have helped. One of > > my 1TB volumes is finally migrating. > > > > WHat’s I’ve noticed is if I allocate a new 1TB volume, I can migrate this > > between NFS and Ceph and it takes only about a 1 minute. I assume this is > > because it’s “thin provisioned” and there’s no actual data on the volume. > > > > But these other volumes I’m trying to move are also “thin provisioned” but > > they’re a part of a LVM group. Does making a thin provisioned device part > > of a LVM group defeat the thin provisioning? I know these volumes weren’t > > full, but I thought perhaps there’s a chance that since it’s a pv in a LVM > > config, that maybe that defeats the thin provisioning and it counts it as a > > full 1TB volume? I’m just spitballing but I’m trying to understand how this > > works so we can do the right thing when provisioning additional volumes. > > > > Also, the behavior I’m seeing is it takes a very long time before I see the > > block image show up on the Ceph side. Perhaps it preallocated a image > > before copying the data? But it seemed strange that I wouldn’t immidiately > > see the image appear on the Ceph side after initiating a migration. It’s > > hard to see what’s actually going on from the logs and the interface. > > > > Thanks > > -jeremy > > > > > > > > > On Tuesday, Jan 16, 2024 at 11:29 PM, Jeremy Hansen > > (mailto:jer...@skidrow.la)> wrote: > > > I changed copy.volume.wait to 72000 > > > > > > But I just noticed: > > > > > > kvm.storage.online.migration.wait and kvm.storage.offline.migration.wait. > > > Worth changing this? > > > > > > Thanks > > > -jeremy > > > > > > > > > > On Tuesday, Jan 16, 2024 at 11:01 PM, Jithin Raju > > > > mailto:jithin.r...@shapeblue.com)> wrote: > > > > Hi Jeremy, > > > > > > > > Have you checked the ‘wait’ parameter? Used as wait * 2 timeout. > > > > > > > > -Jithin > > > > > > > > From: Jeremy Hansen > > > > Date: Wednesday, 17 January 2024 at 12:14 PM > > > > To: users@cloudstack.apache.org > > > > Subject: Re: Issues migrating primary storage > > > > Unfortunately the upgrade didn’t help: > > > > > > > > Resource [StoragePool:3] is unreachable: Volume > > > > [{"name”:”bigdisk","uuid":"8f24b8a6-229a-4311-9ddc-d6c6acb89aca"}] > > > > migration failed due to > > > > [com.cloud.utils.exception.CloudRuntimeException: Failed to copy > > > > /mnt/11cd19d0-f207-3d01-880f-8d01d4b15020/8f24b8a6-229a-4311-9ddc-d6c6acb89aca > > > > to 5837f4e6-9307-43a9-a50c-8c9c885f25e8.qcow2]. > > > > > > > > > > > > > > > > Anything else I can try? I’m trying to move away from NFS completely. > > > > > > > > -jeremy > > > > > > > > > > > > > > > > > > > > > > > > > > > > On Tuesday, Jan 16, 2024 at 7:06 AM, Suresh Kumar Anaparti > > > > mailto:sureshkumar.anapa...@gmail.com>> > > > > wrote: > > > > Hi Jeremy, > > > > > > > > Can you extend with the config 'migratewait' and check. > > > > > > > > Regards, > > > > Suresh > > > > > > > > On Tue, Jan 16, 2024 at 1:45 PM Jeremy Hansen > > > > > > > > wrote: > > > > > > > > > > > > I have some large volumes I’m trying to migrate from NFS to Ceph/RBD. > > > > 1TB > > > > volumes. These inevitably times out. I extended these configs: > > > > > > > > copy.volume.wait=72000 > > > > job.cancel.threshold.minutes=480 > > > > job.expire.minutes=1440 > > > > > > > > This helped with smaller volumes but large once still eventually fail. > > > > > > > > 2024-01-16 07:50:25,929 DEBUG [c.c.a.t.Request] > > > > (AgentManager-Handler-8:null) (logid:) Seq 1-5583619113009291196: > > > > Processing: { Ans: , MgmtId: 20558852646968, via: 1, Ver: v1, Flags: 10, > > > > [{"org.apache.cloudstack.storage.command.CopyCmdAnswer":{"result":"false","details":"com.cloud.utils.exception.CloudRuntimeException: > > > > Failed to copy > > > > /mnt/11cd19d0-f207-3d01-880f-8d01d4b15020/861a6692-e746-4401-9cda-bd791b7d3b5e > > > > to > > > > b7acadc8-34a1-4d7a-8040-26368dafc21d.qcow2","wait":"0","bypassHostMaintenance":"false"}}] > > > > } > > > > > > > > 2024-01-16 07:50:26,698 DEBUG [c.c.s.VolumeApiServiceImpl] > > > > (Work-Job-Executor-41:ctx-e5baf6dc job-1175/job-1176 ctx-bc7b188b) > > > >