Open nick218 opened 6 years ago
Thanks Nick for your post. Interesting that you were able to get the free cluster with couchdb to work - I got a similar joinchannel error when attempting to install the free cluster with couchdb.
The most reliable fix I came up with was checking out tag 1.0 after doing the initial clone. This is at best a short term fix as it installs Fabric v1.0 instead of v1.1, but at least the install finished successfully.
git clone https://github.com/IBM-Blockchain/ibm-container-service
git checkout tags/1.0
The same issue. In the peer log I see following:
2018-07-12 07:59:22.622 UTC [gossip/comm] authenticateRemotePeer -> DEBU 2ae Authenticated 172.21.169.204:30110
2018-07-12 07:59:22.622 UTC [gossip/comm] createConnection -> DEBU 2af Exiting
2018-07-12 07:59:22.622 UTC [gossip/comm] authenticateRemotePeer -> DEBU 2b0 Authenticated 10.77.155.103:48214
2018-07-12 07:59:22.622 UTC [gossip/comm] GossipStream -> DEBU 2b1 Servicing 10.77.155.103:48214
2018-07-12 07:59:22.622 UTC [gossip/comm] sendToEndpoint -> DEBU 2b2 Exiting
2018-07-12 07:59:22.623 UTC [gossip/comm] readFromStream -> DEBU 2b3 EOF Got error, aborting: %!v(MISSING)
2018-07-12 07:59:22.623 UTC [gossip/comm] func2 -> DEBU 2b4 Client 10.77.155.103:48214 disconnected
2018-07-12 07:59:22.623 UTC [gossip/comm] writeToStream -> DEBU 2b5 Closing writing to stream
2018-07-12 07:59:22.645 UTC [gossip/discovery] periodicalSendAlive -> DEBU 2b6 Sleeping 5s
2018-07-12 07:59:27.647 UTC [gossip/discovery] periodicalSendAlive -> DEBU 2b7 Sleeping 5s
2018-07-12 07:59:32.593 UTC [gossip/discovery] periodicalReconnectToDead -> DEBU 2b8 Sleeping 25s
2018-07-12 07:59:32.648 UTC [gossip/discovery] periodicalSendAlive -> DEBU 2b9 Sleeping 5s
2018-07-12 07:59:37.649 UTC [gossip/discovery] periodicalSendAlive -> DEBU 2ba Sleeping 5s
2018-07-12 07:59:42.650 UTC [gossip/discovery] periodicalSendAlive -> DEBU 2bb Sleeping 5s
2018-07-12 07:59:47.652 UTC [gossip/discovery] periodicalSendAlive -> DEBU 2bc Sleeping 5s
The problem is most probably with the couchdb service. kubectl get svc
will help. Check if the public IP is allocated to the couchdb service or is it in pending state?
Probably I will ask a stupid question but this is what kubectl get svc returns: kubectl get svc
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
blockchain-ca NodePort 172.21.198.1 <none> 30054:30054/TCP 1h
blockchain-couchdb1 NodePort 172.21.169.117 <none> 30984:30984/TCP 1h
blockchain-couchdb2 NodePort 172.21.80.124 <none> 31984:31984/TCP 1h
blockchain-orderer NodePort 172.21.214.123 <none> 31010:31010/TCP 1h
blockchain-org1peer1 NodePort 172.21.160.23 <none> 30110:30110/TCP,30111:30111/TCP 1h
blockchain-org2peer1 NodePort 172.21.114.170 <none> 30210:30210/TCP,30211:30211/TCP 1h
External IP is none for all services. Does it tell you guys anything useful?
@amontero0611 If you are using paid scripts, the services are created using this file. As the type of service there is loadbalancer, you will get an external-IP for each service in the list. The fact that the external-IP is none means something is wrong.
Thank you @dhyey20 . I am not using paid scripts, just following the instructions here. This procedure worked fine with the previous version. As I urgently need a hyperledger for prototyping, I've installed the old version as @donbr said and it just works fine. I wonder if someone succeeded installing the new version with the published scripts.
I have the same issue, going through instructions like @amontero0611 . Did anyone find a solution for this problem?
@amontero0611 we have not got a chance to look at this issue yet. If you need a hyperledger-fabric urgently, feel free to get one here
Thank you very much. As I mentioned previously in this thread, I followed what @donbr said and managed to deploy a blockchain previous version. That one works fine.
Salvo indicado de otro modo más arriba / Unless stated otherwise above: International Business Machines, S.A. Santa Hortensia, 26-28, 28002 Madrid Registro Mercantil de Madrid; Folio 1; Tomo 1525; Hoja M-28146 CIF A28-010791
@amontero0611 could you tell me in few steps how did you managed to do it with previous version? when i put in console:
git clone https://github.com/IBM-Blockchain/ibm-container-service git checkout tags/1.0
i only get error message. Could you give me some hints how to get ibm-container-sevice v.1.0? I would be very thankful
I had to update the configs for the paid cluster to get it to deploy correctly. The changes were merged in recently and update this file: cs-offerings/kube-configs/blockchain-couchdb-services-paid.yaml
You have to clone the ibm-container-service repository and checkout the tags from inside the ibm-container-service directory. It should work.
Saludos / Regards,
Ángel L. Montero Sánchez Arquitecto de Industria - Gobierno
amontero@es.ibm.com @angelmonterosn
IBM España Sta. Hortensia 26-28 Madrid 28002 913976611
From: Andrzej Bryk notifications@github.com To: IBM-Blockchain/ibm-container-service ibm-container-service@noreply.github.com Cc: amontero0611 amontero@es.ibm.com, Mention mention@noreply.github.com Date: 25-07-18 08:12 PM Subject: Re: [IBM-Blockchain/ibm-container-service] joinchannel fails during completion (paid) with couchdb (#107)
@amontero0611 could you tell me in few steps how did you managed to do it with previous version? when i put in console: git clone https://github.com/IBM-Blockchain/ibm-container-service git checkout tags/1.0 i only get error message. Could you give me some hints how to get ibm-container-sevice v.1.0? I would be very thankful ? You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or mute the thread.
Salvo indicado de otro modo más arriba / Unless stated otherwise above: International Business Machines, S.A. Santa Hortensia, 26-28, 28002 Madrid Registro Mercantil de Madrid; Folio 1; Tomo 1525; Hoja M-28146 CIF A28-010791
Any updates for scripts v1.1?
Using the script for setting up a paid cluster with couchdb, the installation get stuck with the message
"Waiting for joinchannel container to be Completed"
the paid cluster without couchdb and the free cluster with couchdb works. Any ideas, where the problem is?