There Are Always Gonna Be Lessons Learnt on a Trip

I actually spent my last two weekends in Surabaya. The first trip was for the Kumon roadshow and convention and the second one was for Grid Computing Workshop. I wanted to talk about the second trip. In that trip, I realized how important Standard Operating Procedure (SOP) was. Here’s the thing.

I arrived at Surabaya on Saturday night and planned to spend the next morning in ITS to check the connection in INHERENT network (the network that consists of major universities in Indonesia). When we tested the connection, we found out that the connection from ITS to the grid portal that is located in UI server was too slow, it’s intolerable. We tracerouted the connection from an IP address in ITS to the UI server and the route showed that we went to many routing tables in universities outside the island before finally arriving at UI server. It’s surely an anomaly because normally the route will go from ITS-UNDIP-DIKTI-UI or ITS-UNBRAW-UGM-ITB-UI, in other words it will not go to routers in universities outside Java. We contacted the PIC of INHERENT in ITS who turned out to be a 6th year student and he was just unbelievably mastering in networking. He examined this anomaly and found out that the router in UNDIP had not been working for two days. The interesting thing here is when we contacted the PIC of INHERENT in UNDIP, he wasn’t even aware that the router in UNDIP had been down for two days and he would then find a way to enter the building where the router is located because the building was being renovated which made it diffucult for anyone to enter. Doh.

So what happened with the other route inside the island? Unfortunately, it turned out that the link that connects UI and ITB was also down. Well, the truth is with the money granted for the project (which was huge by the way), we were supposedly able to buy the expensive fiber optic to be used for the links instead of relied on a vendor that won the tender which turned out to let one of the links down in a critical moment with no one to be contacted to post the complaint. Judging by the situation, we decided to not use the INHERENT network and use a telkom speedy (ISP used in ITS for everyday connection) instead because there was no way to do a workshop with that super low connection speed. What I regret the most is it took us the whole day to figure out what the problem that caused the low connection speed really is and finding ways to solve that problem. There was no SOP that specifies what to do if a problem like this occurs including who to contact and the lack of monitoring of the infrastructure by PICs in some cases. I admit that it was also part of our mistake (UI team) which is conducting a workshop without telling all the PICs to make sure the infrastructure is working. But then again, the idea of telling all the PICs was just exhausting, there should be a SOP for this and even if there is already a SOP, it should be reinforced, for the continuity of the project’s sake.

Using inherent network is analogous to using a highway, supposedly smoother and faster. But even if we didn’t use the highway during the workshop, still the workshop was surprisingly a success, with the audience that was more responsive and enthusiastic, the cluster that consisted of temporarily only 5 nodes which worked real good, and the portal that didn’t crash and that was the most important thing. It was my first time to ever speak on a workshop and I was so thrilled by the opportunity. It surely gave me self-satisfaction and confidence. Hell yes, Tika is right. The experience of answering people’s questions is unimaginably satisfying. And I just wanna thank all the people who made this all possible. Like I said, there are always gonna be lessons learned on a trip.

I’m ‘home’ now. Like, for good, if you know what I mean :D