Score:0

Deploying Ceph with RGW Ingress uses [::]:Port instead of virtual ip

br flag

I'm deploying regular Ceph Pacific using cephadm, including RGW. Now I want to use the ingress-service, which includes specifying a virtual ip - so far so good.

Unfortunately this service now creates a haproxy.cfg-file with a binding to "*:80" (used port 80 for regular HTTP access), though I configured virtual ip 10.90.90.2.

As there will be several IPs with several services binding to port 80 I'm not sure whether I'm at fault here (missing a configuration option) or there is an oversight at cephadm, which forces binding to '*' due to the conditional here not being configurable in the ingress service (ingress.py#L132).

Essentially: why is cephadm forcing a bind for "*:<Frontend Port>" on the RGW ingress service, or how can I force this service to bind to specified virtual ip? I'm not entirely sure that editing the configured haproxy.cfg will survive a reboot of the host, or an orch migration between hosts.

mangohost

Post an answer

Most people don’t grasp that asking a lot of questions unlocks learning and improves interpersonal bonding. In Alison’s studies, for example, though people could accurately recall how many questions had been asked in their conversations, they didn’t intuit the link between questions and liking. Across four studies, in which participants were engaged in conversations themselves or read transcripts of others’ conversations, people tended not to realize that question asking would influence—or had influenced—the level of amity between the conversationalists.